Dosimetric comparison and biological evaluation of PET- and CT-based target delineation for LA-NSCLC using auto-planning

Abstract

Using auto-planning, the dosimetric and biological differences between PET- and CT-based target delineation in LA-NSCLC were studied. Twenty-three patients with IIIA-IIIB NSCLC were included in this retrospective study. For each patient, two AP plans (Plan_PET, Plan_CT) were generated based on PET- and CT-based gross tumor volume (GTVPET, GTVCT). The volume, boundary and center of mass (COM) of GTVPET and GTVCT were compared. Dosimetric indicators such as mean lung dose (MLD) and so on were evaluated. Tumor control probability (TCP) of GTVPET and GTVCT and normal tissue complication probability (NTCP) of total lung and heart were calculated. A paired-samples t-test was used to check for significant differences (p < 0.05) between dataset. Volume of GTVPET was significantly smaller than that of GTVCT. Under the premise that GTVPET met the clinical requirements in Plan_PET, GTVCT couldn't satisfy the requirements. GTVCT met the clinical requirements in Plan_CT, and four cases of GTVPET could not satisfy the requirements. Compared with Plan_CT, Plan_PET significantly reduced MLD, V5, V10, V13, V15, V20, V30 and V40 of total lung, and MHD, V30 and V40 of heart, and MUs. No significant difference was observed with respect to Dmax of spinal cord. TCP of GTVPET in Plan_PET was significantly higher than that of GTVCT. NTCP of total lung in Plan_PET was significantly lower than that in Plan_CT. There were differences in volume, boundary, and COM of targets based on the two delineation methods. These led to differences in dosimetric and biological indicators. For LA-NSCLC, the way that most hospitals only use CT to delineate the target should be careful consideration.