Validation Of MRI, CT Or FDG-PETt Based Gross Tumor Volume Using Cross-Sectional Autopsy Specimen In Rabbits With Head And Neck Cancer

Part one:Establishment Head and Neck Model of Rabbit Using VX2Cell and Investigation of Its Growth Pattern and Image FeatureObjects:To establish rabbit VX2transplantation head and neck cancer animal models, to understand its growth pattern, observe images feature of the tumor in CT, MRI and18F-FDG-PET/CT scans.Materials and Methods:Purebred male and female New Zealand white rabbits (n=10), weighing2-3kg were provided by the Animal Center of Nanfang Hospital. VX2tumor cells were obtained through intramuscular injection in the hind leg of rabbits with tumor in vivo. Rabbits with tumor were provided by the Zhongshan Medical University. All animalsreceived care in compliance with the Principles of Laboratory Animal Care formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. All experimental procedures were approved by the Care of Experimental Animals Committee of Laboratory Animal Center, Nanfang Hospital. The rabbits were anesthetized with intramuscular injection of3%pentobarbital. Solid tumors in the groin area of the rabbits with tumor were stripped off under anesthesia and sterile surgical conditions. The samples were rinsed with normal saline and the capsule was removed. Tissue with rapid growth was collected from the edge of the tumor and was cut into lmm3blocks using a scalpel. The samples was placed in Hank’s solution and centrifuged at800rpm in a centrifugal scrubber for4minutes. Then the tissue blocks were placed in a200-300mesh sieve homogenizer and homogenized with appropriate amount of0.9%normal saline. Suspended cells were collected and centrifuged in a centrifugal scrubber at800r/min for5minutes until the supernatant was clear. Serum-free DMEM medium was added to make a cell suspension with a diluted concentration of1×107cells/ml. The number of viable cells were counted using the trypan blue exclusion test, the percentage of viable cells was>95%. For tumor cell implantation, experimental rabbits were anesthetized and fixed in the supine position. Then under CT positioning guide, VX2tumor cell suspension was injected to head and neck of rabbit with0.5ml. The experimental animals underwent CT scanning1,2and3weeks after inoculation, when model successful. Therefore, the growth character and image feature were observed by CT, MRI and F-FDG-PET/CT in addition to the specimen obtained by naked eyes and the results from pathology.Results:1. In10rabbits, modeling successful in8after2-3weeks of VX2cell line inoculation. The success rate is80%. But2cases were dead in accident,6rabbits were brought into the study.2. Th head and neck cancer VX2tumor grew rapidly, invasive growth, and more than3-4weeks later, necrosis would be occured gradually from the center to outward of the tumor.3. The pathological biopsy of head and neck tumor was in accord with poorly differentiated squamous cell carcinoma.4. It can be clear displayed in CT, MRI and18F-FDG-PET/CT images about nasal pharyngeal structures and tissue surrounding tumor in10cases of rabbit with VX2transplantation head and neck cancer.5. In CT scan imaging, we found the VX2tumor in rabbit mainly infringe the deep muscles make pharynx side muscle bundle to swell, the density increases. And muscle fat interval changed to fuzzy, even disappear. Bone structure damage can be occured in someone. CT contrast enhancement showed that the contrast enhancement of tumor is weak and irregular. CT contrast enhancement characteristics of VX2tumor are similar to head and neck cancer.6. In MR scan imaging, the characteristic of VX2tumor is equisignal on T1WI or slight high signal on T2WI, and it is lower than normal nasopharyngeal mucosa signal. It is high signal on T1WI MRI contrast enhancement phase, enhanced obviously.7. In18F-FDG-PET/CT imaging, it shows that radioactive material is high concentration in VX2tumor. The maximum SUVmax and minimum SUVmax are6.90g/ml and4.96g/ml in region of interest, respectively.Conclusions:1. The rabbit VX2transplanted carcinoma of head and neck cancer model was established with high successful rate.2. Its disease progress of rabbit VX2head and neck cancer (HNC) model is resemble to the human HNC.18F-FDG-PET/CT, MRI and CT can clearly display the structure of nasopathrynx and surrounding organs.So it is an ideal animal model to study human HNC disease. Part two:Validation of MRI or CT Based Gross Tumor Volume Using Cross-Sectional Autopsy Specimen in Rabbits with Head and Neck CancerObjects:To validate the gross tumor volume (GTV) delineation based on magnetic resonance imaging (MRI) or computed tomography (CT) with cross-sectional autopsy in rabbits bearing head and neck cancer.Materials and Methods:Ten head and neck cancer models were built in New Zealand rabbits with transplantation of VX2cell line. Modeling was successful in six rabbits. Each rabbit received MRI scan and CT scan in the same position and immobilization was sacrificed and fixed by cryopreservation within24hours. Then, all cryopreservated rabbits underwent cross-sectional autopsy in the same position with imaging scan with sectional thickness of4.0mm. And to get the cross-sectional autopsy images by digital single-lens reflex camera. After three-dimensional coregistration, the GTV was delineated based on CT, MRI and cross-sectional autopsy images, respectively. Then, the volume value of GTV, overlap GTV and volume difference ratio (VDR) based on different imaging scans and cross-sectional autopsy specimens were computed by treatment planning system.Results:1. Differences between GTVMRI, GTVCT and GTVSA:In6rabbits bearing head and neck cancer, no significant difference (P=0.943) was observed between the average value of GTVs delineated on MRI (8.20±2.56cm3), CT (8.40+2.20cm3), and cross-sectional autopsy images (8.11+2.88cm3).2. Difference between GTVMRI&sA and GTVCT&SA:No significant difference (P=0.099) was observed between the average value of overlap GTVs delineated on MRI (6.63±1.93cm3) and CT (5.67±1.93cm3) with cross-sectional autopsy images. 3. Difference between VDRMRI-SA and VDRCT-SA:Significant difference (P=0.001) was found between the average value of VDRMRI-SA (0.180±0.060) and VDRCT-SA (0.309±0.091).Conclusions:1. VDR is a useful parameter in evaluating the difference in tumors/targets.It is much better than volume parameter alone.2. The accuracy of GTV definition based-on MRI is better than that based-on CT in rabbits bearing head and neck cancer. Part three:Validation of Three Segmentation Tools for18F-FDG-PET/CT Based Gross Tumor Volume Using Cross-Sectional Autopsy Specimen in Rabbits with Head and Neck CancerObjects:To validate the gross tumor volume (GTV) delineation based on three segmentation tools for F-FDG-PET/CT with cross-sectional autopsy in rabbits bearing head and neck cancer.Materials and Methods:Ten head and neck cancer models were built in New Zealand rabbits with transplantation of VX2cell line. Modeling was successful in six rabbits. Each rabbit received F-FDG-PET/CT scan in the same position and immobilization was sacrificed and fixed by cryopreservation within24hours. Then, all cryopreservated rabbits underwent cross-sectional autopsy in the same position with imaging scan with sectional thickness of4.0mm. And to get the cross-sectional autopsy images by digital single-lens reflex camera. After three-dimensional coregistration, the GTV was delineated based on CT, MRI and cross-sectional autopsy images, respectively. Then, the volume value of GTV, overlap GTV and volume difference ratio (VDR) based on different imaging scans and cross-sectional autopsy specimens were computed by treatment planning system.Results:1. Differences between GTVvis, GTV50%, GTV2.5and GTVSA:In6rabbits bearing head and neck cancer, significant difference (P=0.005) was observed between the average value of GTVs delineated on visual (9.76±2.23cm3),50%SUVmax (8.67±2.49cm3), SUV=2.5(10.11±3.35cm3), and cross-sectional autopsy images (8.11±2.88cm3), of which GTV2.5(P=0.009) and GTVvis(P=0.035) are larger than GTVSA, respectively.2. Differences between GTVviS&sA、GTV50%&SA and GTV2.5&SA:No significant difference (P=0.284) was observed between the average value of overlap GTVs delineated on visual(6.90±1.47cm3),50%SUVmax(6.81±1.80cm3) and SUV=2.5(6.58±1.97cm3) with cross-sectional autopsy images.3. Differences between VDRvis-sAVDR50%-sA and VDR2.5-SA:Significant difference (P=0.042) was found between the average value of VDRvis-SA (0.216±0.073), VDR50%-SA (0.178±0.055) and VDR2.5-sa(0.270±0.037), of which VDR50%-SA is better than VDR2.5-SA(P=0.041). No significances were detected between each of the other two groups.Conclusions:1.18F-FDG-PET/CT can involve the display lesions within the tumor heterogeneity of distribution changed.2. The accuracy of GTV definition based-on50%SUVmax is better than that based-on SUV=2.5in rabbits with head and neck cancer. Part four:Validation of gross tumor volume delineation based on MRI, CT or FDG-PET/CT with cross-sectional autopsy in rabbits bearing head and neck cancerObjects:To compare F-18-fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET)/computed tomography (CT), magnetic resonance imaging (MRI) and CT alone based tumor volume measurements in rabbits bearing head and neck cancer (HNC) with cross-sectional autopsy and to validate the gross tumor volume (GTV) delineation based on above three imaging methods.Materials and Methods:Ten head and neck cancer models were built in New Zealand rabbits with transplantation of VX2cell line. Modeling was successful in six rabbits. Each rabbit received CT, MRI or F-FDG-PET/CT scan in the same position and immobilization was sacrificed and fixed by cryopreservation within24hours. Then, all cryopreservated rabbits underwent cross-sectional autopsy in the same position with imaging scan with sectional thickness of4.0mm. And to get the cross-sectional autopsy images by digital single-lens reflex camera. After three-dimensional coregistration, the GTV was delineated based on CT, MRI and cross-sectional autopsy images, respectively. Then, the volume value of GTV, overlap GTV and volume difference ratio (VDR) based on different imaging scans and cross-sectional autopsy specimens were computed by treatment planning system.In addition,50%SUVmax was defined as a better standard of GTV definition in18F-FDG-PET/CT.Results:1. Differences between GTVCT, GTVMRI,GTVPET and GTVSA:No significant difference (P=0.873) was observed between the average value of GTVs delineated on CT (8.40±2.20cm3), MRI (8.20±2.56cm3), FDG-PET (8.67±2.49cm3) and cross-sectional autopsy images (8.11±2.88cm3).2. Differences between GTVCT&SA、GTVMRI&SA and GTVPET&SA:Significant difference (P=0.028) was observed between the average value of overlap GTVs delineated on CT (5.67±1.93cm3), MRI (6.63±1.93cm3) and FDG-PET(6.81±1.80cm3) with cross-sectional autopsy images, of which GTVCT&SA is smaller than GTVPET&SA (P=0.039). No significances were detected between each of the other three groups.3. Differences between VDRCT-SA, VDRMRI-SA and VDRPET&SA:Significant difference (P=0.008) was found between the average value of VDRCT-SA(0.309±0.091), VDRMRI-SA (0.180±0.060) and VDRPET&SA(0.178±0.055), of which VDRPET is better than VDRCT (P=0.017), VDRmMRI is better than VDRCT (P=0.019). No significance was detected between VDRmMRI-SA and VDRPET-SA.Conclusions:1.18F-FDG-PET/CT or MRI was found to be more accurate modalities than CT in GTV definition of head and neck cancer.2. The GTV delineation of FDG-PET/CT in head and neck cancer is not superior to MRI.