Impact of three threshold segmentation of 18FDG PET image on target volume delineation and radiotherapy treatment planning of non-small cell lung cancer / 中华放射肿瘤学杂志

ABSTRACT

Objective To evaluate the accuracy, of different threshold segmentation of 18FDG PET for target volume delineation of non-small cell lung cancer(NSCLC) and the potential influence on radiotherapy treatment planning. Methods Eight NSCLC patients who had tumor with clear margin on CT scan and the amplitude of tumor movements not more than 5 mm were enrolled. PET scans were carried out at 1 h after intravenous injection of 18FDG with CT image for attenuation revisement. Gross target volume (GTV) delineated on CT image ( GTVCT ) was used as the standard. Then, GTVs were delineated on PET image with three different threshold segmentation of 42% Imax(total) (42% of maximum voxel intensity within the tumor) ,Iback + 20% Imax-back(max) (mean background intensity + 20% of normalized background-subtracted maximum voxel intensity within the tumor) and Iback -20% Imax-back(slice) (mean background intensity + 20% of normalized background-subtracted maximum voxel intensity of each slice within the tumor) ,the corresponding GTV was named as GTV42%, GTV20%max and GTV20%slice. Both the size of GTV42%, GTV20%max, GTV20%slice and GTVCT,and the coverage over GTVCT for each GTV were compared. A three dimensional margin of 1 cm were added to GTVCT, GTV42%, GTV20%max and GTV20%slice to form corresponding PTVCT, PTV42%, PTV20%max and PTV20%slic e. Three dimensional conformal radiotherapy treatment plans were designed based on PTVCT,PTV42% , PTV20%max and PTV20%slice respectively for each patient. The prescription dose of all PTVs was 66 Gy in 33 fractions in 6.6 weeks. Both the volume accepting dose less than 95% of prescription dose within PTVCT ( VPTV ) and the lung V20 were compared among the four plans based on different PTVs. Tumor control probability(TCP) as well as lung normal tissue complication probability (NTCP) were also compared. Resuits Eight patients were enrolled in this study. Median deviation of volume between GTVPET and GTVCT were -54.1% , -21.5 % and 5.3 % for GTV42% , GTV20%max and GTV20%slice, respectively. Median coverage over GTVCT of GTV42% , GTV20%max and GTV20%slice was 45.9% ,78.0% and 95.3% respectively( F = 57.50,P＜0.01). Median 7.5% of VPTV was observed for radiotherapy treatment plan based on PTV42% ,which meant that it might induce median 1% decrease of TCP comparing with that of radiotherapy treatment plan based on PTVCT. Whereas,there were only 1.3% and 0.0% of VPTV for treatment plans based on PTV20%max and PTV20%slice respectively. As far as TCP was concemed, both PTV20%max group and PTV20%slice group were superior to PTV42% group,there was no significant difference among PTV20%max group, PTV20%slice group and PTVCT group. Lung V20 and lung NTCP showed no significant difference among all groups. Conclusions The threshold segmentation of Iback + 20% Imax-back(slice) , being slice specialized, might be an optimal threshold segmentation for target volume delineation of lung caner. Independent of information of target volume provided by CT scan in advance,it is recommended to use for the target volume delineation of NSCLC with atelectasis.