Tumoural perfusion as measured by dynamic computed tomography in head and neck carcinoma.

PURPOSE To investigate the intra- and interobserver variability, as well as the intra- and interpatient variability of CT-determined tumour perfusion in head and neck tumours, and to evaluate the preliminary value of this parameter as predictive factor of local failure after treatment by definitive radiotherapy. MATERIALS AND METHODS In 41 patients the perfusion of a primary head and neck squamous cell carcinoma was estimated using dynamic CT. A 40-ml intravenous bolus of a low-osmolar non-ionic contrast agent was rapidly injected over 5 s (8 ml/s), while a dynamic acquisition of image data was obtained during the first pass at the level of the largest axial tumour surface. A time-density curve was constructed for the primary tumour and the carotid artery. The perfusion in the selected tumour region of interest was calculated by dividing the slope of the tumour-time density curve by the maximal value in arterial density. Tumour volume was calculated on the CT-images and correlated with perfusion rate. RESULTS The mean perfusion rate was 86.4 ml/min per 100 g (median, 80.6; SD, 43.05; range, 31.7-239.8 ml/min per 100 g). No systematic difference was found between the measurements performed by two independent observers. The intratumoural COV was 0.22, the intertumoural COV 0.37. No correlation was found with tumour volume. Ten out of 20 patients with a perfusion rate < 80 ml/min per 100 g were not locally controlled, while nine out of 21 patients with a value > 80 ml/min per 100 g did show a local failure (P = 0.19). CONCLUSIONS CT-determined perfusion measurements of head and neck tumours are feasible. No correlation with tumour volume and a sufficiently large COV were found to consider this parameter as a possible prognostic factor for outcome after radiotherapy. More patients need to be investigated to test the hypothesis that tumours with a low CT determined perfusion rate have a higher risk of local failure.