Cinzia Iotti, Presenter: Nothing to Disclose

Nunzia D'Abbiero, Abstract Co-Author: Nothing to Disclose

Federica Fioroni, Abstract Co-Author: Nothing to Disclose

Elisa Grassi, Abstract Co-Author: Nothing to Disclose

Marta Paiusco, Abstract Co-Author: Nothing to Disclose

Tamara Palmieri, Abstract Co-Author: Nothing to Disclose

Ramundo Dafne, Abstract Co-Author: Nothing to Disclose

Antonino Romeo, Abstract Co-Author: Nothing to Disclose

Roberto Sghedoni, Abstract Co-Author: Nothing to Disclose

Annibale Versari, Abstract Co-Author: Nothing to Disclose

This retrospective study tries to correlate the H&N tumor volumes obtained with a visual evaluation, performed by an experienced nuclear medicine physician (NM), with a pure activity-based quantitative method carried out by a radiation oncologist(RO)

18 pts with H&N cancer enrolled for IMRT have undergone CT-FDG PET study. An experienced NM performed the PET images segmentation with Advantage Sim - GE. No fixed threshold value was applied, since the tumor volume has been drawn as corresponding to the most correct expression of the disease extension by a visual analysis, after a preliminary SUV data evaluation. Then, the RO has run an automatic segmentation of the same images with Eclipse -Varian. The maximum activity value has been measured for every site of disease; then, a threshold corresponding to a percentage of this value has been used to segment the tumor. The relative threshold has been obtained by phantom measurements in different conditions of uptake volume and source-background ratio (mean value 42% of the max uptake ± 3SD). In our practice, this threshold-based contour has been applied to delineate the definitive gross tumor volume (GTV).

A linear proportion between SUV and activity has been observed. With the NM's approach, the percentage of the max SUV values enveloping the tumor volume ranged from 25% to 40% (mean 30%). These values are significantly lower than the ones used by the RO. As expected, the NM volumes showed a good agreement with the contour set by the RO with respect to the shape of the disease but, with respect to the dimension, they were found to be greater by a mean factor of 20% (range 0-43).

The contribution of an experienced NM is advised to define the best possible extension of the disease. Nevertheless, the values derived from this evaluation are out of the limits reported by the published studies and could reveal a too predominant subjective assessment. A more confident and objective method of tumor segmentation, possibly common to the RO and to NM physician, is needed

A more confident and objective method of tumor segmentation, possibly common to the RO and to NM physician, is needed

Iotti, C, D'Abbiero, N, Fioroni, F, Grassi, E, Paiusco, M, Palmieri, T, Dafne, R, Romeo, A, Sghedoni, R, Versari, A, The FDG-PET for the H&N Cancer IMRT Planning: The Nuclear Medicine Physician and the Radiation Oncologist Approaches for Defining the Tumor Volume—A Retrospective Analysis of 18 Patients Treated at Reggio Emilia Hospital.  Radiological Society of North America 2007 Scientific Assembly and Annual Meeting, November 25 - November 30, 2007 ,Chicago IL. http://archive.rsna.org/2007/5015266.html