| Part1. The detection of background SUV of lung, aorta and liver in patients with NSCLCObject:The aim of this study was to investigate the variation of background SUV in normal tissues in patients with NSCLC.Material and Methods:85patients who previously untreated, biopsy-confirmed Stage I to III NSCLC and were performed whole-body PET CT scan in Medical Nuclear Center of Shanghai Cancer center between May2010and April2012were eligible. In order to obtain the region of interest (ROI), various sites in normal tissues, including left upper lobe (LUL), left lower lobe (LLL), right upper lobe (RUL), right middle lobe (RML), right lower lobe (RLL), aorta (AOR), liver left zone (LILZ), liver median zone (LIMZ) and liver right zone (LIRZ), were auto-contoured with True D software on work station in PET CT center. The database of maximum SUV (SUV max), average SUV (SUV ave), tumor-to-background ratio (TBR) and Coefficient of Variance (CV) were acquired, and were analyzed with STATA11version.Results:In normal tissue of lung, the mean SUV max and SUVave of LUL, LLL, RUL, RML and RLL was0.71±0.25and0.40±0.13,0.83±0.38and0.48±0.19,0.75±0.42and0.41±0.19,0.77±0.42and0.45±0.24,0.99±0.51and0.57±0.31, respectively. In terms of SUV max or SUVave, differences was significantly observed in the group of LUL vs LLL (P=0.0128and0.0011), the same as LUL vs RLL (P=0.000and0.000), LLL vs RLL (P=0.0217and0.0255), RUL vs RLL (P=0.0011and0.0001) and RML vs RLL (P=0.0030and0.0069), respectively.Conclusions:The vast variation of SUV, not only among normal tissues, but also in different sites of organ was observed. The delineation of tumor volume based Source-To-Background Ratio is one of the most widely used methods in NSCLC. Obviously, it is arbitrary to select the others normal tissue as the background in lung cancer, such as liver and aorta. Part2. The impact of the simulation of respiration movement on delineation of target volume with PET CT in phantomObject:The aim of this study was to investigate the impact of the simulation of respiration movement on delineation of target volume with PET CT in phantom.Material and Methods:NEMA ICE-body Phantom Set with6spheres volume of0.52ml、1.15ml、2.6ml、5.5ml、11.4ml and26.5ml were injected the solution of0.48-3.2uCi/ml18F-FDG to simulate different source background ratio. The phantom was placed on the home-made three dimensions movement apparatus and were performed on a PET CT scanner (Siemens Biograph16HR, Siemens Medical Solutions) with a field of view of58.5cm in diameter and an axial length of15.5cm. PET data was acquired by using an acquisition time of2.5min per bed. The volume of these spheres were obtained with various auto-contouring methods in MIM5.6.6and analyzed with STATA11version.Results:In our study, the SBR method without adding the tumor movement factor shown the best correlation coefficient of0.969(p<0.00) in dynamic group. However, the50%SUVmax method have a correlation coefficient with0.971(p<0.00) which was the best in static group.Conclusions:The variation inter-delineation methods was observed in our study, and the SBR method without adding the movement factor is consistent with the clinical situation, when contouring the target volume with PET CT in NSCLC. Part3. The impact of variation of organs’background SUV on delineation of tumor volume in PET CT with NSCLCObject:To determine how to contour the target volume with PET CT is more reasonable according to the variability of background SUV of lung, aorta (blood pool) and liver and contoured the gross tumor volume (GTV) with different methods in patients with non-small-cell lung cancer (NSCLC).Methods:Thirty-seven patients who were previously untreated, pathologically confirmed NSCLC and performed PET CT scan were enrolled into this study. To obtain normal18F-fluoro-deoxy-glucose (18F-FDG) uptake value within normal tissues, various sites, including lung left upper lobe(LUL), lung left lower lobe(LLL), lung right upper lobe(RUL), lung right middle lobe(RML), lung right lower lobe(RLL), aorta(AOR), liver left zone(LILZ), liver median zone(LIMZ) and liver right zone(LIRZ), in normal tissues were measured. The GTV of PET-based delineated (GTVPET) by the method of absolute SUV2.5, the40%and50%percentage of SUV maximal intensity level and source-background ratio (SBR). The CT-based GTV (GTVCT) was considered the standard to which all PET-based gross tumor volumes were compared. The correlation coefficient was analyzed to compare gross tumor volume by various delineation methods. Linear and logarithmic regression analyses were used to determine the relationship between the volume of the GTVCT and GTVPET.Results:Lung showed significant lower SUVs and less stability than aorta, followed by liver. The maximum SUV (SUVmax) and average SUV (SUVave) of lung, aorta and liver were0.82±0.32,2.35±0.37,3.24±0.50, and0.49±0.18,1.68±0.32,2.34±0.36, respectively. The SUVs of lung varied from lobe to lobe. Twelve methods were used to contour the GTV. The median volume of GTVCT、. GTV2.5、GTV4o%、GTV50%、GTVTLA、 GTVLLA、GTVAA、GTVLA、GTVTLM、GTVLLM、GTVAM and GTVLM were16.28.11.82、8.01、5.47、16.22、16.69、10.58、9.05、14.24、9.05and7.36cm3. The correlation coeffient of GTVCT to GTV2.5、GTV40%、GTV50%、GTVTLA、GTVLLA、GTVAA、GTVLA、 GTVTLM、GTVLLM、GTVAM and GTVLM were0.788.0.66.0.628.0.776.0.813.0.791.0.775.0.803.0.804.0.784and0.763. The method SBR used the SUVave of lung lobe with tumor as background showed the best correlation with the volume of CT-based GTV (r=0.813).Conclusions:Our results have shown that the volume of tumor based on SBR method correlated well with CT-based tumor volume. It would be more reasonable and precise to contour tumor volume according to background of lung lobes (GTVLLA) with tumor when using SBR to delineate GTV for patients with NSCLC. |
PDF Full Text Request