| Objective To compare the positional and volumetric differences of internal gross target volumes (IGTV) and planning target volumes (PTV) delineated based on four-dimensional CT (4DCT) images and 18F-fluorodeoxyglucose positron emission tomography CT (PET-CT) for the primary tumor of non-small cell lung cancer (NSCLC).Methods Patients with NSCLC underwent enhanced three-dimensional (3D) CT scan followed by enhanced 4DCT scan of the thorax under normal free breathing and then 18F-FDG PET-CT scan was performed based on the same position and positioning parameters. The gross target volumes (GTVs) of ten phases of 4DCT were contoured and IGTV10 was obtained by fusion of the ten GTVs. IGTVMIP of the primary tumor were contoured on the MIP images of 4DCT. IGTVPET were determined with different threshold methods using the auto-contouring function and manual contouring. The differences in the position, volume, degree of inclusion (DI) and conformity index (CI) of the targets between IGTVPET and IGTV10 as well as IGTVMIP were compared. IGTVPET15% was defined on PET-CT images using a fixed threshold of 15% of the maximum standard uptake values (SUVmax). The PTV4D and PTVPET were obtained by adding a 10 mm margin around IGTV10 and IGTVPET15%, respectively. The differences in target volume, position and coverage between PTV4D and PTVPET were evaluated.Results For ten NSCLC patients, only in superior-inferior (SI) directions, there were significant differences between the centric coordinate parameters of IGTVPEt (IGTVPEti.5 excluded) and IGTV10 (z=-2.703- -2.293, P<0.05). In volume terms, IGTVPET2.0 approximated closely to IGTV10 with median volume ratios of 0.90, followed by IGTVPET20% to IGTV10 with median volume ratios of 0.84, but there was no significant difference between them (P>0.05). The DI of IGTVPET1.5 in IGTV10 was higher than the DIs of other IGTVPET in IGTV10 (z=-2.803- -2.429, P<0.05). The CI of IGTVPET2.0 and IGTV10 was superior than the CIs of IGTVPET35% and IGTV10, IGTVPET40% and IGTV10, IGTVPET2.5 and IGTV10, IGTVPET3.0 vs IGTV10 (z=-2.803- -2.191, P<0.05).For ten NSCLC patients, there were significant differences between the centric coordinate positions of IGTVPET and IGTVMIP only in SI directions (P=0.014-0.044). In volume terms, IGTVPET2.0 and IGTVPET20% approximated closely to IGTVMIP with mean volume ratios of 1.02 and 1.06 (P=0.806), respectively. The highest CI were between IGTVMIP and IGTVPET2.0 or between IGTVMIP and IGTVPET20%(0.46 and 0.45, P=0.603). The best DI were IGTVMIP in IGTVPET20% or IGTVMIP in IGTVPET2.0 (0.61 and 0.61, P=0.963).For fifteen NSCLC patients, there was no significant difference of centroid position between PTVPET and PTV4D (P=0.589,0.147,0.096). There was no statistical difference between PTVPET and PTV4D(P=0.156). However, PET-CT resulted in an alteration of PTV exceeding 20% in 5 out of 15 patients. In 10 patients, the PTVPEt volume increased (average of 30%). For five patients the PTVPET volume decreased (average of 11%). The mean DI of PTV4D in PTVPEt was 85%. The percentage of PTVPEt that was not included within the PTV4D ranged from 7% to 46%. The DI between PTVPET and PTV4D showed no significant correlation to the 3D motion vector (P=0.134,0.405).Conclusions The IGTVPET contoured by SUV≥2.0 or 20% of the maximal SUV approximate closely to IGTV10 as well as IGTVMIP in target size and the extent of spatial missing is relatively lower between them, but neither of them could replace IGTV10 as well as IGTVMIP in spatial position. When the primary tumor target of NSCLC was delineated basing on PET-CT, the suitable SUV threshold should be selected meanwhile the target position correction should done referring to 4DCT images.Although there were no significant differences in centroid position and target volumes between PTVPET and PTV4D, the spatial mismatch was apparent between them and the mismatch showed no correlation to the tumor displacement. |
PDF Full Text Request