Predictive and prognostic role of functional imaging of head and neck squamous cell carcinomas

Predicting radiotherapy (RT) treatment response and eventual locoregional disease control is an important component of the ongoing effort to improve the therapeutic ratio in the management of head and neck squamous cell carcinomas. The development of clinically useful predictive and prognostic imaging biomarkers has been limited by significant tumor heterogeneity in both the tumor and its microenvironment. Various advanced imaging techniques have been evaluated in the head and neck squamous cell carcinoma patient, which now offer a strategy to identify and quantify this heterogeneity, characterizing the tumor at baseline and its response to RT. The most promising of these techniques include dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), DCE computed tomography, diffusion-weighted MRI, and (18)F-fluoromisonidazole positron emission tomography (PET) all relying on the spatiotemporal quantification of a contrast agent within a region of interest that can be further analyzed by various pharmacokinetic models. Despite the small study populations, several consistent observations have been reported that warrant further validation. Features associated with a favorable RT response include tumors with an effective vasculature characterized by rapid and high influx rates of the contrast agent and its effective clearance with little or no regions of hypoxia. (18)F-deoxyglucose-PET imaging remains an active area of investigation with the metabolic tumor volume parameter appearing to offer potential predictive value. Characterizing changes during a course of RT may offer greater predictive value. Both DCE-MRI and diffusion-weighted MRI can identify physiological changes within the first 1-2 weeks of treatment that are correlated with long-term clinical outcome. Identifying persistent hypoxia with (18)F-fluoromisonidazole-PET during a course of RT suggests an increased risk of relapse. Whether this is due to an inability to favorably remodel the tumor's vasculature has not been clearly demonstrated to date. Future research goals include the need to further validate these promising imaging biomarkers especially in larger cohorts of patients, characterizing the optimal threshold cutoffs and to refine the predictive value by incorporating the assessments of early tumor responses to therapy that offer the potential for increased specificity because it reflects the biological stress responses.