【作者】 管一晖；

【导师】 林祥通； 戴嘉中； 潘力； 盛晓芳； 汪洋；

【作者基本信息】 复旦大学 ， 影像医学与核医学， 2008， 博士

【副题名】脑肿瘤~（11）C-Choline PET/CT显像及其在放射治疗计划中的应用

【摘要】 一、选题的目的与设计思路:脑肿瘤按来源通常分为源自颅内各组织的原发性肿瘤和由身体它处转移至脑内的转移性肿瘤两大类。原发性脑肿瘤发病率居全身恶性肿瘤第11位,原发性脑肿瘤以胶质瘤为主,尽管目前各种综合治疗方法不断进展,胶质瘤患者的中位生存期仍十分有限,尤其是病理分级高的肿瘤;而转移性脑肿瘤的发病率更高,在常见肿瘤中有20-40%患者会出现脑转移。目前脑部计算机断层扫描（CT）和核磁共振成像（MRI）及其相应的增强显像是用于诊断脑部肿瘤的主要常规方法,近年来发展的影像诊断技术正电子发射断层显像（PET）被用于脑肿瘤的良恶性鉴别、术前病理分级、病程分期、鉴别肿瘤复发或坏死、探测残留肿瘤、立体定向穿刺、放疗计划的制定、判断肿瘤对治疗的敏感性、患者预后的判断等方面。PET以各种放射性示踪剂作为显像的基础,最常用的放射性示踪剂为经典的2-[¹⁸F]fluoro-2-deoxy-D-glucose（FDG）,为临床提供了CT、MRI尚难以给予的各种关于肿瘤能量代谢的生物学信息,尤其在肿瘤复发与残留肿瘤的鉴别等方面显得尤其重要。近年来肿瘤生物靶区体积（BTV）概念的出现,预示着PET的应用将会更为广泛,它通过在肿瘤组织的血流灌注、代谢、增殖活性、乏氧、肿瘤相关受体、血管生成及凋亡等方面的显示为放射治疗的进一步优化提供了重要技术平台。但从技术上看,仅仅FDG显然是不够的,由于大脑皮质对于葡萄糖的相对高摄取,使得FDG对于脑肿瘤的显像特异性及对低度恶性脑肿瘤显像的敏感性受到较大限制,因此PET尚需要其他的显像剂如氨基酸类显像剂、胆碱类、乙酸类及神经受体显像剂、乏氧显像剂、嘧啶类等显像技术的从多种不同代谢途径反映了肿瘤的异质性,提供更好的诊断特异性及对肿瘤形态精确描绘,为PET在脑肿瘤方面的应用提供了更好的技术平台。本研究拟通过探讨¹¹C标记药物PET成像的技术因素评价、PET在脑肿瘤临床诊断的可靠性、对放射治疗方案制定的实验基础技术探索,¹¹C标记药物对脑肿瘤诊断价值评价,对脑肿瘤坏死与复发的鉴别及在肿瘤放疗实施过程的研究等方面对PET在脑肿瘤方面的应用进行研究。二、研究结果:第一部分¹¹C-PET显像的模型实验研究本实验研究分四个实验部分进行。1、实验通过应用PET/CT仪器系统模型与Hoffman 3D脑模型进行不同正电子核素成像测试,以热区、冷区分辨率和线性及均匀度等指标和HOFFMAN 3D脑模型的测试显示图进行比较,结果显示¹¹C和¹⁸F两种正电子核素在图像分辨能力方面无显著性差异,其中¹¹C的图像与¹⁸F图像相比在小的热区分辨能力显示方面还是有一定的畸变,而¹⁸F与¹¹C的测试结果显示无论热区和冷区大小测量时均会出现少许高估和低估的显像,需要在临床应用中测量病灶大小时加以注意。2、实验自行建立了在Hoffman 3D脑模型中,制作模拟脑肿瘤病灶的方法,模拟肿瘤的大小为5mm和15mm,该方法可用于PET/CT及MRI显像。实验显示MRI对病灶大小估计准确,而在PET显像中,模拟肿瘤的球体内外预置的放射性浓度比（1∶5和1∶10）对热区大小测量有影响,高摄取的病灶大小可能会有一定的被高估可能,提示在临床显像中应加以注意。3、实验把上述脑肿瘤模型应用于两类不同的放射治疗系统（Varian clinical600-C放射治疗系统和Accuray的Cyber knife）中,模型图像可以通过光盘传输进入放疗计划系统（TPS）,按要求的方式进行调整处理,采用图像融合的方法勾画靶区,执行放疗计划,说明PET图像在放疗计划系统进行靶区设定是可行的,可以进一步用于临床患者的生物靶区的勾画,实现生物靶区放射治疗。4、实验通过8例胶质瘤患者的脑PET显像结合MRI及病理结果进行研究,可见¹¹C-Choline PET显像显示病灶范围最大,涵盖了肿瘤亚临床的浸润部分,与实际临床制定肿瘤治疗范围时所需目标体积最为相符;MRI显像的结果与肿瘤实体部分的体积最为接近;¹⁸F-FDG PET显像的结果总体小于肿瘤实体部分的体积但与MRI显像结果在统计学上无显著差异,说明¹¹C-Choline较为适于放射治疗靶区勾画,值得进一步积累更多病例深入研究。第二部分¹¹C-Choline与¹⁸F-FDG联合显像对脑瘤诊断价值本部分研究通过178例脑肿瘤的¹¹C-Choline PET、¹⁸F-FDG及MRI的显像研究,结果表明与MRI比较,MRI诊断脑肿瘤的灵敏度及特异性高于¹⁸F-FDG PET;¹¹C-Choline PET显像灵敏度与特异性与MRI相似,其差异无统计学意义,MRI在解剖结构的显示方面好于PET/CT,但在肿瘤生物学行为和肿瘤的活力显示方面PET优于MRI,¹¹C-Choline PET对病灶定性最为准确,在肿瘤的分级方面有一定的帮助,且可以鉴别放射性坏死与复发。第三部分PET/CT在脑肿瘤坏死与复发鉴别中的研究本部分研究通过55例脑肿瘤放疗后怀疑复发或放射性坏死患者进行¹¹C-CholinePET、¹⁸F-FDG及MRI的显像研究,结果显示¹¹C-Choline PET显像可提高脑肿瘤坏死及复发的诊断准确率,研究表明¹¹C-Choline PET/CT显像诊断准确率为90.9%,显著高于MRI（85.5%）及¹⁸F-FDG PET/CT（72.7%）显像。然而¹¹C-Choline PET/CT显像在鉴别脑肿瘤坏死及复发中也存在一定的假阳性及假阴性,需要在临床诊断中结合病史和系列增强MRI的显像表现,才能得到正确的结果。第四部分PET/CT在脑肿瘤放射治疗定位中的应用价值研究本研究应用SIEMENS公司出品的BIOGRAPH SENSATION 16型PET/CT及BIOGRAPH 64 HD型PET/CT、Varian公司的clinical 600c和Nomos公司的PEACOCK TPS系统进行调强适形放疗技术设计研究。通过对44例脑肿瘤患者进行定位¹¹C-Choline PET/CT显像,应用激光线进行定位,将采集的PET/CT图像通过光盘存储,送至TPS进行治疗计划,由放疗物理师和医师根据PET和CT两组资料应用计划系统的图像融合软件,使图像达到融合标准后制定放疗计划靶区。治疗期间详细记录患者症状及早期放射反应,放射治疗结束后3个月再进行近期疗效评价。结果建立了非¹⁸F-FDG PET/CT和PEACOCK适形调强放疗系统实用融合图像方法学;44例受试患者中共有28（63.64%）例患者的治疗计划通过PET/CT显像发生了改变,主要是治疗的范围（PTV）增加（14/44,31.82%）,部分患者（14/44,31.82%）经过PET/CT融合后,所制定的PTV范围减小;44例患者经过PET/CT技术定位放射治疗后均取得良好疗效。建立在适形调强放疗系统的PET/CT融合图像方法学是我们完成治疗的首要问题,这种方法的应用将提高对生物靶区体积制定的精确性,使临床放疗后患者缓解的可能增加。三、本研究论文可以得出以下结论:1.¹¹C和¹⁸F两种正电子核素在图像分辨能力方面无显著性差异,其中¹¹C PET图像与¹⁸F图像相比在小的热区分辨能力显示方面还是有一定的畸变,而¹⁸F与¹¹C的测试结果显示无论热区和冷区大小测量时均会出现少许高估和低估的显像;在PET显像中,不同的放射性浓度会对肿块大小测量产生影响,高摄取的病灶,大小可能会被高估;PET图像在放疗计划系统进行靶区设定是可行的;¹¹C-Choline PET显像显示病灶范围实际临床制定肿瘤治疗范围时所需目标体积最为相符,可以作为放射治疗生物靶区勾画。2.MRI及其增强扫描是目前脑肿瘤的常规应用手段,PET/CT是其有益的补充。MRI在解剖结构的显示方面好于PET/CT,¹¹C-Choline PET显像在灵敏度与特异性方面与MRI相似,但在肿瘤生物学行为和肿瘤的活力显示方面PET优于MRI,¹¹C-Choline PET肿瘤的分级方面有一定的帮助,且可以鉴别放射性坏死与复发。3.PET/CT在鉴别脑肿瘤放射性损伤和复发方面具有重要意义,其中¹¹C-CholinePET/CT诊断准确率为90.9%,显著高于MRI及¹⁸F-FDG PET/CT显像。但同时也存在一定的假阳性及假阴性,需要在临床诊断中结合病史和系列增强MRI的显像表现,才能得到正确的结果。4.在脑胶质瘤放疗中,CT、MRI以及PET等图像提供了可以互补的有利于放疗计划制定的信息,PET/CT可以作为生物靶区制定的重要指标,有助于提高肿瘤靶区的精确确定和治疗疗效监测。在脑胶质瘤放疗计划制定中,由于受PET显像空间分辨率的限制,脑本底的影响等,制定放疗计划的过程需要多种图像共同参与。基于¹¹C-胆碱、¹⁸F-FLT、¹¹C-乙酸、¹⁸F-FDG等多种PET示踪剂的联合应用,结合常规CT、MRI显像以及临床资料的综合信息的诊断和治疗可能是将来PET在脑肿瘤放疗发展的最终方向,尚宜今后继续探索。更多还原

【Abstract】 Purpose and Project design:Brain tumors can be classified by origin into primary or metastatic. Primary brain tumor has a high occurrence ranking the 11^th in all malignant tumors.The most common type of primary brain tumor is glioma.Despite improvements in various treatment strategies,median survival of patients is limited,especially the high grade tumor;the metastatic brain tumor has a even higher occurrence with a possibility of 20-40%in all tumor patients.Computed tomography（CT） and magnetic resonance imaging（MRI） with their enhanced imaging are now the routine methods of brain tumor diagnosis.With the development of positron emission tomography（PET）, it has been applied in various ways of brain tumor imaging including malignancy differentiation,pre-operation grading,staging, post-treatment monitoring,stereotaxis needle biopsy,radiation treatment planning,prognosis prediction,etc.With the foundation of various radiopharmaceuticals,especially the classic 2-[¹⁸F]fluoro-2-deoxy-D-glucose（FDG）,PET can provide the biological information of tumor metabolism that neither CT nor MRI can.The biological tumor volume （BTV） is a newly developed concept including the blood infusion, metabolism,proliferation activity,tumor specific receptor,vascular generation,etc.of the tumor.To define such BTV,FDG only is far from enough.Due to the high uptake of FDG in cerebral cortex,brain tumor imaging with FDG is limited in specificity and sensitivity of low grade tumor.Other radiopharmaceuticals like amino-acid,choline,acetate,etc. are needed to provide more accurate diagnosis and delineation of the tumor.This study intended to evaluate the imaging quality of 11C labeled tracers,their liability in brain tumor diagnosis and the application in radiation treatment planning of brain tumor. Results:PartⅠStudy of ¹¹C-PET imagingThis part of study includes 4 experiments.Experiment 1:PET/CT imaging of Hoffman 3D brain model using multiple positron radiopharmaceuticals.Comparing the results in hot region,cold region,linearity and homogeneity,our study showed no significant difference in image resolution between ¹¹C- and ¹⁸F- labeled tracers. However,in small hot regions there is certain aberration in resolution test with ¹¹C-labeled images compared to ¹⁸F-labeled images and all results showed mild deviation in both hot and cold region,which should be taken into consideration in clinical use.Experiment 2:I established brain tumor imaging model in Hoffman 3D brain model.The sizes of the tumor models were 5mm and 15mm in diameter.MRI showed the actual size of the tumor while in PET imaging,different background concentration of the radiopharmaceuticals（1:5 and 1:10） has certain impact in accessing of tumor size.Tumors with high uptake were likely to be over-accessed,which should be taken into consideration in clinical use.Experiment 3:I applied the tumor model in two different radiation treatment systems（clinical 600-C of Varian and Cyber knife of Accuray）. Images of the model were transferred through compact disc into the treatment planning system（TPS） and adjusted to meet the requirement. Target delineation was performed with fused images.My study showed that radiation treatment planning using PET imaging is feasible and can be applied into clinical use.Experiment 4:8 patients with glioma were studied using PET imaging and MRI with contrast.Our study indicated that ¹¹C-Choline PET showed the largest area of the tumor including the sub-clinical infiltrating area which consist mostly with the actual treatment planning area;the result of MRI was closest to the solid part of the tumor;result of ¹⁸F-FDG PET was smaller than the solid part of the tumor but there is no statistical significance compared to the result of MRI.In conclusion,our study showed that ¹¹C-Choline PET was the most suitable imaging technology among the three methods in radiation treatment planning of brain tumors. PartⅡThe diagnostic value of ¹¹C-Choline combined with ¹⁸F-FDG PET imagingI studied 178 cases of brain tumor with MRI,¹¹C-Choline PET,¹⁸F-FDG PET imaging.Out study showed that the diagnostic sensitivity and specificity of MRI was higher than ¹⁸F-FDG PET but similar as ¹¹C-Choline PET.MRI is superior in showing the anatomical structure but inferior in revealing the biological activities.¹¹C-Choline PET is the most accurate method among all three in locating the tumor;it also showed value in tumor grading and differentiating necrosis from recurrence.PartⅢPET/CT in differentiating necrosis from recurrence of brain tumorsI studied 55 cases of post-radiation brain tumor with MRI,¹¹C-Choline PET, ¹⁸F-FDG PET imaging.My study showed that ¹¹C-Choline PET could improve the diagnostic accuracy in differentiating necrosis from recurrence.The accuracy of ¹¹C-Choline PET was 90.9%,which was significantly higher than ¹⁸F-FDG PET and MRI.However,there was still certain rate of false positive and false negative and clinical information and MRI with contrast have great importance in diagnosis.PartⅣThe application of PET/CT in radiation treatment planning of brain tumorsUsing Biograph Sensation 16 and Biograph 64 HD PET/CT scanner of Siemens, clinical 600c of Varian and PEACOCK TPS of Nomos,44 patients with brains tumors were studied with 11C-Choline PET.Metal radio-active points（18F） and laser beams were used for alignment.The acquired data was transferred to TPS through compact disc.The target delineation was performed by radiation therapists and nuclear medicine physicians based on fused images.Patients were closely monitored for symptoms and early reaction of radiation and were accessed 3 months later.I established the methodology of incorporating non-FDG PET/CT imaging in PEACOCK radiation treatment planning.Among all 44 patients,28（63.64%） patients significantly changed their treatment plan due to PET/CT,14 （31.82%）patients increased the target volume and 14（31.82%） decreased. All patients were treated successfully.The application of PET/CT will increase the accuracy of biological target volume planning and increase the possible of clinical remission.Conclusion:1.There is no significant difference in resolution between ¹¹C-labeled and ¹⁸F-labeled tracers.In small hot regions there is certain aberration in resolution test with ¹¹C-labeled images compared to ¹⁸F-labeled images and all results showed mild deviation in both hot and cold region. Different background concentration of the radiopharmaceuticals has certain impact in accessing of tumor size.Tumors with high uptake were likely to be over-accessed.Radiation treatment planning using PET imaging is feasible.¹¹C-Choline PET showed most consistent result with the target volume which could be used in delineating biological target volume.2.MRI with contrast is the conventional method currently in brain tumor imaging,PET/CT plays a compensate role.MRI is superior in showing the anatomical structure but inferior in revealing the biological activities. ¹¹C-Choline PET is the most accurate method among all three in locating the tumor;it also showed value in tumor grading and differentiating necrosis from recurrence.3.PET/CT has important value in differentiating post-radiation necrosis from recurrence of brain tumors.The diagnostic accuracy of ¹¹C-Choline PET/CT was 90.9%,which was significantly higher than MRI and ¹⁸F-FDG PET/CT.There was still certain rate of false positive and false negative and clinical information and MRI with contrast have great importance in diagnosis.4.In radiation treatment of gliomas,CT,MRI and PET imaging provided compensative information in treatment planning.PET/CT can be applied to delineating biological target volume that improves the planning accuracy and therapeutic effect.Limited by the spatial resolution,background uptake,etc.,treatment planning requires multiple methods of imaging. The combined use of multiple radiopharmaceuticals including ¹¹C-Choline, ¹¹C-acetate,¹⁸F-FDG,with conventional CT,MRI and clinical information leads to the future of diagnosis and treatment with PET in brain tumors.更多还原