| Part OneComparison of 11C-CHO and 18F-FDG PET in a Rabbit Lung VX2Tumor Model and Correlation with MVD and PCNAObjective: To establish a stable implanted rabbit lung VX2 tumor model. Therabbits had positron emission tomography (PET) studies with 11C-choline (CHO) and18F-2-fluoro-2-deoxy-D-glucose (FDG), and maximal standardized uptake values(SUVmax) were calculated. Immunohistochemistry was undertaken with tumormarkers of angiogenesis and proliferative activity. To determine and compare thesuitability of CHO and FDG PET for reflecting tumors angiogenesis and proliferativeactivity.Methods and materials: Fifty-four New Zealand White rabbits which weighted2.5～3kg were used in the experiment. Under general anesthesia, a needle wastransthoracically inserted into the right lung, 0.5 ml of viable VX2 tumor cellsuspension was slowly injected through the needle. Plain CT scan was performed toevaluate growth of the tumor one week after implantation. The appearance of a solitary pulmonary nodular was considered to be the sign of successful tumorimplantation.CHO and FDG PET were performed 10～11 days after implantation. One ear veinwas cannulated for administration of CHO and FDG. CHO PET was performed 5minutes after intravenously injection of 1mCi CHO. FDG PET was performed 2hafter CHO administration and 1h after 0.5mCi FDG administration. PET wasperformed by use of a GE Discovery LS PET/CT scanner. The data were transmittedto Xeleris workstation. CT, PET and PET/CT fusion images were obtained, andtumors SUVmax were calculated. The animals were euthanized after examination.Their lungs were fixed with 10％formalin, transaxially sliced, embedded in paraffin,sectioned with a microtome. The sections were stained with hematoxylin and eosin,and immunostained for CD34 and proliferating cell nuclear antigen (PCNA).Assessment of cellular density, microvessel density (MVD) and PCNA index wasperformed by computer-assisted image analysis. The relationship between CHOSUVmax and FDG SUVmax with tumor size, cellular density, MVD and PCNA indexwere statistically analyzed.Results: Of the 54 rabbits, 36 had a solitary pulmonary VX2 tumor, the rate ofsuccessful generation of a solitary VX2 tumor was 66.7％. But 2 rabbits died afterPET examination and 1 rabbit with large necrotic area in tumor center were notincluded, 33 rabbits were enrolled in this study at last. Of the remaining 18 rabbitswhich were excluded from the study, 8 showed pleural dissemination and 2 formeddiffuse bilateral pulmonary metastases, and 8 died after implantation.All lung VX2 tumors differently accumulated CHO and FDG. The mean CHOSUVmax was 4.02±3.07 (1.4～12.2), and the mean FDG SUVmax was 5.70±3.45(1.0～13.0). The mean size of tumor was 1.68±1.61cm3 (0.13～8cm3). In a high-powerfield(200×, 0.739mm2), the mean cellular density was 547.364±64.78(413～708), themean MVD was 35.784±13.63 (13～64), the mean PCNA index was 42.34％±15.26％(3.23％～75.87％). FDG SUVmax was higher than CHO SUVmax (P=0.003), and they were positively correlated (r=0.578, P=0.000). CHO SUVmax significantly andpositively correlated with PCNA index (r=0.786, P=0.000), rather than tumor size,cellular density and MVD. FDG SUVmax significantly and positively related to MVD(r=0.525, P=0.002) and PCNA index (r=0.531, P=0.001).There was a marginallypositive correlation between FDG SUVmax and tumor size (r=0.335, P=0.057), Nosignificant correlation was found between FDG SUVmax and cellular density(r=-0.105, P=0.561). Finally, a weakly positive correlation was found between MVDand tumor size (r=0.362, P=0.038).Conclusions: In a rabbit lung VX2 tumor model, both CHO SUVmax and FDGSUVmax correlated with PCNA index. Both CHO and FDG PET could reflect tumorproliferative activity. The correlation between CHO SUVmax and PCNA index wasmore obvious than FDG SUVmax, and CHO PET could more effectively reflect tumorproliferative activity than FDG PET. FDG SUVmax correlated with MVD, and FDGPET could reflect tumor angiogenesis. CHO SUVmax did not statistically correlatewith MVD, and CHO PET could not reflect tumor angiogenesis.Part TwoThe Value of Perfusion CT in reflecting tumor angiogenesis andComparison with CHO and FDG PET in a Rabbit ModelObjective: Perfusion CT was performed in a rabbit lung VX2 tumor model inorder to determine the suitability of perfusion CT for reflecting tumor angiogenesis.And the correlation between perfusion CT and CHO PET and FDG PET wasstatistically analysed.Methods and materials: The experimental animals, PET examination andimmunohistochemistry were identical to part one. Perfusion CT was performed using GE Discovery Lightspeed Qx/4i multi-slice spiral scanner after PET examination. Atotal of 5ml contrast agent (Omnipaque, 300) was injected through a cannulated earvein by a power injector at a rate of o.5 ml/s. Technique parameters were as follows:cine mode, 80kV, 80mA, ls/rotation, 5mm sections, and matrix 512×512, FOV 12cm.The total scanning time was 50s, and scanning started 5s before intravenous injection.Perfusion CT parameters including blood flow (BF), blood volume (BV), mean transittime (MTT) and permeability surface (PS) were obtained using GE Perfusion 3software from tumor, and corresponding functional maps were reconstructed.Statistical analysis was performed between perfusion CT parameters and MVD andCHO SUVmax and FDG SUVmax.Results: The results of immunostaining for MVD and PET were identical to Partone. The time-density curve (TDC) of rabbit lung VX2 tumors demonstrated quicklyrise and quickly fall, and then maintained a high platform. The time to peakenhancement of tumor was slightly later than the aorta. Tumors could be demonstratedclearly on functional maps. The mean BF was 80.55±67.16 (6.57～262.24)ml/100mg/min, mean BV was 5.73±4.13 (0.68～16.68)ml/100mg, mean MTT was7.92±5.22 (2.31～22.15) s, and mean PS was 11.16±8.82 (0.39～33.08)ml/100mg/min.MVD significantly and positively correlated with BF (r=0.736, P=0.000) and BV(r=0.609, P=0.000), and negatively correlated with MTT (r=-0.505, P=0.003), ratherthan PS (r=0.089, P=-0.624). Tumor size was positively related to BV (r=0.354,P=0.043). A significantly positive correlation was found between FDG SUVmax andBF (r=0.786, P=0.000) and BV (r=0.529, P=0.002), a negative correlation was foundbetween FDG SUVmax and MTT (r=-0.505, P=0.003). No correlation could be foundbetween FDG SUVmax and PS (r=0.077, P=0.669). There were no correlationbetween CHO SUVmax and BF (r=0.296, P=0.095) and BV (r=0.145, P=0.420) andMTT (r=-0.196, P=0.273) and PS (r=-0.063, P=0.727).Conclusions: In a rabbit lung VX2 tumor model, a significantly positivecorrelation was found between MVD and BF or BV, a reverse correlation was found between MVD and MTT, but no correlation between MVD and PS could be found.Perfusion CT can effectively reflect tumor MVD and further indirectly reflect tumorangiogenesis. Although the correlation between FDG SUVmax and MVD was slightlyinferior to perfusion CT parameters, FDG PET can reflect tumor angiogenesisreplacing perfusion CT to a certain extent.Part ThreeThe Clinical Value of 11C-CHO and 18F-FDG PET in DifferentiatingPulmonary Malignant Tumors from Benign LesionsPurpose: PET with 18F-FDG is used for diagnosis of lung cancer but has somelimits; more specific PET radiophannaceuticals would be welcome. 11C-labeled CHOis a new radiopharmaceutical potentially useful for tumor imaging, since it isincorporated into cell membranes as phosphatidylcholine and can reflect cellproliferative activity. The aim of this study was to investigate whether it C-CHO PEThas advantages over 18F-FDG PET in differentiating pulmonary malignant tumorsfrom benign lesions.Methods and materials: We retrospectively studied the CHO PET and FDG PETimages of 36 patients who had suspected pulmonary malignant lesions on CT. In thesepatients, 22 were males, 14 were females, their average age was 64.5±11.2 years andthe age range was 38-82 years. Twenty-two patients had malignant tumors, includingadenocarcinomas (n=18), squamous cell carcinomas (n=2), bronchioloalveolarcarcinoma (n=1) and small cell carcinoma (n=1). Fourteen patients had benignlesions, including chronic inflammation (n=6), tuberculosis (n=3), benign tumors (n=4) and inflammatory pseudotumor (n=1). All patients fasted for at least 8 h before the PET studies. Blood glucose levelswere checked to ensure that they were less than 8mmol/L. CHO was injectedintravenously, and whole-body scanning was started after 5 min. FDG was injectedintravenously 1h after CHO administration, and FDG PET scanning started 1h afterFDG administration. Scanning from the base of skull to the middle of the thigh wasaccepted using the GE Discovery LS PET/CT system. First, low-dose CT wasperformed, with 140 kV, 90 mA, 0.8 s per CT rotation, a pitch of 6, a table speed of22.5 mm/s, and no specific breath-holding instruction. Immediately afterward, a PETemission scan was obtained without changing the patient's position. PET images werereconstructed with CT attenuation correction. CT, PET and PET/CT fusion imageswere obtained on a Xeleris workstation equipped with fusion software. Forsemiquantitative analysis, maximal standardized uptake values (SUVmax) werecalculated. The tumor size, CHO SUVmax, FDG SUVmax and FDG SUVmax/CHOSUVmax were analysed in both malignant tumors and benign lesions. The benignlesions were further classed into chronic inflammation, tuberculosis and benigntumors. CHO SUVmax, FDG SUVmax and FDG SUVmax/CHO SUVmax inmalignant tumors and benign lesions and the three benign subgroups were comparedwith each other. ROC curve analysis was performed in order to compare the value ofCHO SUVmax, FDG SUVmax and FDG SUVmax/CHO SUVmax for differentiationof pulmonary tumors.Results: The mean size was 82.58±183.41cm3 (1～777.48cm3) in 22 malignanttumors, and 45.74±77.32cm3 (0.125～216cm3) in 14 benign lesions. CHO uptake in 22malignant tumors were significantly lower than FDG uptake (P=0.000). CHOSUVmax ranged from 1.3 to 9.0, and the mean was 4.34±2.22. FDG SUVmax rangedfrom 2 to 41.3, and the mean was 12.92±9.75. A positive correlation was foundbetween CHO SUVmax and FDG SUVmax (r=0.597, P=0.003). No correlation couldbe found between the tumor size and CHO SUVmax (r=-0.049, P=0.827) and FDGSUVmax (r=0.018, P=0.936) in 22 malignant tumors. CHO SUVmax in 14 benign lesions ranged from 0.6 to 6.8, and the mean was3.51±2.10. FDG SUVmax ranged from 0.3 to 26, and the mean was 6.76±7.51.Although CHO SUVmax was lower than FDG SUVmax, but the difference was notstatistically significant (P=0.274, P=0.052). A positive correlation was also foundbetween CHO SUVmax and FDG SUVmax (r=0.809, P=0.000). There wassignificantly positive correlation between the tumor size and CHO SUVmax (r=0.715,P=0.004) and FDG SUVmax (r=0.850, P=0.000) in 14 benign lesions. The mean FDGSUVmax/CHO SUVmax was 3.18±1.90 (0.76～6.89) in malignant tumors, and1.91±1.84 (0.12～7.29) in benign lesions. A marginally significant difference could befound between them (P=0.056).In benign tumors, the mean CHO SUVmax was 1.55±1.00 and the mean FDGSUVmax was 1.00±0.39. Both of them were much lower than malignant tumors(P=0.013, P=0.012). The mean CHO SUVmax and FDG SUVmax of chronicinflammation were 3.46±1.59 and 4.91±2.59, respectively. Both of them were lowerthan malignant tumors, but only the difference of FDG SUVmax is statisticallysignificant (P=0.031). The mean CHO SUVmax and FDG SUVmax of tuberculosiswere respectively 6.27±0.68 and 18.73±7.25. Although both of them were slightlymore than malignant tumors, which was not statistically significant (P=0.119,P=0.257). Moreover, CHO SUVmax and FDG SUVmax of tuberculosis were higherthan benign tumors (P=0.003, P=0.008) and chronic inflammation (P=0.045,P=0.020). No significant difference could be found in CHO SUVmax (P=0.130) andFDG SUVmax (P=0.451) between benign tumors and chronic inflammation.The mean FDG SUVmax/CHO SUVmax of pulmonary benign tumors was0.90±0.61 (0.26～1.5), which was lower than malignant tumors (P=0.033). The meanFDG SUVmax/CHO SUVmax of chronic inflammation was 2.03±2.37 (0.12～7.29),and the mean FDG SUVmax/CHO SUVmax of tuberculosis was 2.95±0.97 (2.09～4.0).Although both of them were slightly lower than malignant tumors, but the differencewas not statistically significant (P=0.171, P=0.884). Moreover, no significant difference could be found in FDG SUVmax/CHO SUVmax among pulmonary benigntumors and chronic inflammation and tuberculosis (P=0.342, P=0.163, P=0.486).ROC analysis showed the area under the curve of CHO SUVmax, FDG SUVmaxand FDG SUVmax/CHO SUVmax were 0.593, 0.763 and 0.740, respectively. Theiroptimal threshold value for diagnosis malignant tumors was respectively 3.1(sensitivity 68.2％, specificity 50％, and accuracy 61.1％), 6.6 (sensitivity 81.8％,specificity 71.4％, and accuracy 77.8％) and 1.8 (sensitivity 72.7％, specificity 71.4％,and accuracy 72.2％).Conclusions: Both CHO and FDG PET could not differentiate pulmonarymalignancy from chronic inflammation and tuberculosis. The value of FDG PET indifferentiating malignant tumors from benign lesions was superior to CHO PET. Thevalue of FDG SUVmax/CHO SUVmax in differentiating malignant tumors frombenign lesions was similar to FDG PET.
|
PDF Full Text Request