【作者】 刘志芳；

【导师】 于金明；

【作者基本信息】 山东大学 ， 临床医学， 2012， 博士

【摘要】 研究背景和研究目的肺癌是全球最常见的恶性肿瘤,已成为癌症死亡的首位病因。非小细胞肺癌(non-small cell lung cancer, NSCLC)约占肺癌总数的75%～80%。随着影像学的发展、早期检出手段的提高和NSCLC有效的全身治疗,NSCLC脑转移的发生率在逐年升高。长期以来,临床诊断脑内转移瘤主要靠CT和MRI,尤其是MRI。但是常规MRI平扫及强化,反映的是病变的含水量、血流的流空现象和造影剂透过血脑屏障进入病灶等情况,不能提供肿瘤增殖程度和异质性等生物学信息。PET即正电子发射计算机断层显像(positron emission tomography, PET),是利用正电子核素标记葡萄糖、氨基酸等人体代谢物作为显像剂,通过病灶对显像剂的摄取来反映其代谢变化,从而为临床提供疾病的生物代谢信息。PET/CT融合了PET的功能显像和CT的精确定位,可快速获得多层面断层影象、三维定量结果以及三维全身扫描,在脑转移瘤早期诊断及预后判断方面有重要意义。近年来,11C-胆碱(11C-choline,11C-CHO)作为一种新的PET示踪剂用于多种恶性肿瘤的显像。在肿瘤细胞内,胆碱的唯一代谢途径是参于磷脂的合成。胆碱通过磷酸化途径进入肿瘤细胞,最终整合到细胞膜上,形成胆碱入胞后的化学捕获机制。并且胆碱一旦被磷酸化后就停留在细胞中,即“化学滞留”,这是胆碱可以用于肿瘤显像的生物学基础。活体组织内磷脂酰胆碱31P磁共振波谱显示：许多肿瘤细胞膜上的磷脂酰单碱成分增多,而相应的正常组织细胞膜上胆碱代谢物浓度很低或无法探测；治疗有效时,治疗后肿瘤细胞膜上的磷酰单碱成分减少。因此,瘤组织的胆碱摄取速率是反映细胞膜合成速率的指标,也是肿瘤细胞增殖的指标。11C-CHO代谢显像剂的优点是肿瘤显像清晰,周围正常组织摄取胆碱的数量非常有限,放射性很低,而肿瘤组织摄取胆碱相对高,肿瘤/非肿瘤放射性比值高。绝大多数脏器在1～5min时摄取率达最高值,之后逐渐降低或处于相对稳定状态。所以注射11C-CHO约5min后即可进行PET显像。虽然在一些良性肿瘤及瘤样病变胆碱摄取亦增高,有一定的假阳性及假阴性率,但如果应用适当,11C-CHO在脑肿瘤显像具有潜在的优势。数十年来,全脑放疗(whole brain radiotherapy, WBRT)在脑转移治疗中保持着主导地位,能够使脑转移患者获得3-6月的生存期。近来多项临床研究报道替莫唑胺(temozolomide, TMZ),尤其是低剂量TMZ联合WBRT对NSCLC(?)(?)转移有一定的疗效。影响脑转移生存的临床变量已经被深入研究,重要的变量包括：性别、年龄、体力状态评分、脑转移灶数目(单个或多个)、原发灶病理类型(鳞癌或非鳞癌)以及颅外肿瘤状况(控制或非控制)。其中,体力状态评分被一致地认为是影响生存的决定性因素,其次是治疗方法。原发疾病确诊至出现脑转移的间隔时间也影响预后,间隔时间长者为预后有利因素。11C-CHO PET在脑转移瘤中的良好显像,且能反映肿瘤细胞增殖情况,可能具有在预测脑转移瘤的生存的价值。研究内容1、11C-胆碱PET/CT对非小细胞肺癌脑转移的诊断价值；2、低剂量替莫唑胺同步全脑放疗对比单纯全脑放疗治疗非小细胞肺癌脑转移；3、影响非小细胞肺癌脑转移生存的多因素研究。研究方法1.对30例临床诊断非小细胞肺癌脑转移或可疑脑转移的患者进行颅脑MRI检查及11C-CHO PET/CT检查,对11C-CHO摄取进行视觉分析和半定量分析,探讨SUV值与性别、年龄、病理类型及ECOG评分的相关性。计算11C-CHO PET/CT显像的各个指标对肺癌脑转移诊断的敏感性和特异性。2.对90例确诊为非小细胞肺癌脑转移的患者随机分组,分别进行低剂量替莫唑胺同步全脑放疗与单纯全脑放疗,并进行长期随访,观察其临床疗效及相应的毒副反应。3.通过Cox比例风险模型探讨性别、年龄、肿瘤病理类型及治疗方式的等各种临床因素及脑内转移灶的11C-CHO摄取等生物因素与生存的关系。4.病理检查：对所取组织行常规病理HE染色,光学显微镜观察。5.统计学处理：使用SPSS13.0软件进行数据分析,定量测定结果用均数±标准差(X±s)或中位数表示,两组样本均数的比较采用t检验,多个不同水平间的比较采用单因素方差分析(One-Way ANOVA)或Mann-Whitney秩和检验,数据间的相关性采用Pearson或Spearman(?)目关性分析。ROC曲线用于比较各个指标的诊断效能。应用Cox比例风险模型进行多变量回归分析。以P<0.05为具有统计学显著意义。研究结果1.30例肺癌患者经手术、磁共振显像及随访证实有26例存在脑转移,共发现45个脑转移瘤灶,CHO PET显示的高中代谢灶39个,诊断脑转移瘤的敏感度性、特异性和准确性分别是86.7%(39/45),特异度85.7%(6/7),准确性86.7%(45/52)。2.11C-CHO显像剂的4个半定量指标SUVmax、L/MCU、L/WM、L/CTX的均值分别是：1.1276、4.1160、9.0769、4.7203,对于脑转移瘤均有良好的诊断效能,所有AUC均>0.9(P<0.01)。3.转移灶长径和SUVmax存在统计学相关性,相关系数r=0.454(P<0.01),肿瘤长径越大CHO摄取值越高；不同性别、不同年龄段间肿瘤大小及SUVmax值均无统计学差异；4.4个半定量指标SUVmax、L/MCU、L/WM和L/CTX鳞癌脑转移中的均值分别是：1.3178、5.1638、10.3739、5.4774,腺癌脑转移均值分别是：1.0007、3.4175、8.2119、4.2156；鉴别不同病理类型的效能均较差,AUC分别是0.642、0.711、0.626、0.619,仅L/MCU有一定的鉴别效能(P=0.018)。5. WBRT联合同步低剂量替莫唑胺对脑内病灶疗效有效率73.3%；单纯WBRT组有效率77.8%(P>0.05)。中位生存期(OS) WBRT联合同步低剂量替莫唑胺组10.0个月,单纯WBRT组7.5个月,联合治疗组有显著生存优势(P=0.002)。6.两组均常见头痛、皮疹及脱发,且脱发多为Ⅲ级以上。单纯WBRT组血液学及胃肠道反应轻微。WBRT联合同步低剂量替莫唑胺组耐受性良好,其最常见的不良反应是血液学毒性和胃肠道反应。两组Ⅱ级以上血液学毒性、胃肠道反应和肝功损伤有统计学差异(P<0.05)。7.对90例患者的生存资料进行Cox比例风险模型分析显示原发灶病理类型、ECOG评分、合并内脏转移及同步替莫唑胺化疗是影响生存的独立因素(P<0.05),而性别、年龄及脑转移灶的数目与生存无关(P>0.05)；对26例脑转移患者的11C-CHO摄取的4个半定量指标与生存的关系进行Cox比例风险模型分析,显示SUVmax、L/WM和L/MCU影响生存(P<0.05),而L/CTX与生存无关(P>0.05)结论1.11-C-CHO PET/CT对NSCLC脑转移有良好的诊断价值,但对于鉴别不同病理分型效能有限。11C-CHO摄取值与性别、年龄无关,而与肿瘤的病理类型及肿瘤大小有关,肺鳞癌脑转移瘤11C-CHO摄取值大于肺腺癌。2. WBRT联合低剂量TMZ治疗NSCLC脑转移较单独WBRT显著延长生存期,且耐受性良好。影响生存的临床因素有原发灶病理类型、ECOG评分、是否合并内脏转移及联合低剂量替莫唑胺化疗；生物因素有脑内转移瘤灶SUVmax、L/WM和L/MCU。这些影响因素可在一定程度上作为预测预后的指标。更多还原

【Abstract】 BACKGROUND AND OBJECTIVE:Lung cancer is the most common cancer worldwide and a leading cause of mortality. Non-small cell lung cancer(NSCLC) accounts for about75%～80%of the total lung cancer. The incidence of brain metastases with NSCLC appears to be rising year by year as a result of superior imaging modalities, earlier detection, and more effective treatment of systemic disease.Over the years, the diagnosis of brain metastases depends mainly on computed tomography (CT) and magnetic resonance imaging (MRI), especially MRI. However, conventional and enhanced MRI scan cannot provide the extent and heterogeneity of tumor proliferation and other biological information because it reflects the water content of the lesion, blood flow void phenomenon and contrast agent through the blood-brain barrier into the lesion. Positron-emission tomography (PET) can provide valuable metabolic information by the use of positron labeled glucose, amino acids and other metabolites in the human body as an imaging agent. PET/CT combines the functional imaging of PET and precise positioning of CT and is used for quick accessing of multi-level images, three-dimensional quantitative results, and three-dimensional body scan, which is important in early diagnosis and prognosis of brain metastases. Recently,11C-choline has been introduced as a new tracer for PET imaging in a variety of malignant tumors. In tumor cells, the only metabolic way of choline is to participate in the synthesis of phospholipid. Choline is incorporated into cells through phosphorylcholine synthesis and is integrated into the cell membranes. Once phosphorylated, choline will stay in the cell, which is called "chemical retention." This provides the biological rationale for the use of [methyl-11C]choline (11C-choline) in oncological PET studies. In vitro phosphorus-31magnetic resonance spectroscopy studies have revealed the high content of phosphorylcholine in most cancers; while in corresponding normal tissues this choline metabolite is present at low or undetectable levels. Moreover, phosphorylcholine will reduce when a treatment is effective. Therefore, the choline uptake rate of tumor tissue reflects not only the rate of synthesis of the cell membrane, but also an indicator of tumor cell proliferation.The tracer of11C-choline has the advantage of clear tumor imaging with the surrounding normal tissue a very limited uptake of choline, a very low radioactivity, a relatively high tumor11C-choline uptake, and a high radioactivity of tumor/non tumor ratio. If applied properly,11C-choline has potential advantages in brain tumor imaging, although that may make a certain false positive and false-negative rate for the reason of an increased uptake in some benign tumors and tumor-like lesions.Over the past several decades, WBRT has been the mainstay of treatment for brain metastases and increased the median survival time to3to6months. More recently, the combination of WBRT with temozolomide (TMZ), especially with a low-dose has shown promising response rates in NSCLC patients with brain metastases. Clinical variables predictive of survival in patients with brain metastases have been well studied. Important variables include:age, performance status (most commonly designated by the Karnofsky performance status [KPS] score, number of brain metastases (single or multiple), primary tumor type, and systemic tumor activity (controlled versus uncontrolled). Of these, the KPS score has consistently been shown to be the major determinant of survival, secondary only to treatment regimen in most studies. Time from primary tumor diagnosis to development of brain metastases holds prognostic value as well, with long intervals being favorable.11C-CHO PET is supposed to have some prognosis value for it good imaging in brain metastases which reflects the proliferation of tumor cells.STUDY CONTENTS:1. Clinical usefulness of11C-choline PET/CT for identifying NSCLC with brain metastasis.2. Low-dose temozolomide concomitant with whole-brain radiotherapy compared with whole-brain radiotherapy alone for non-small cell lung cancer patients with brain metastases.3. Analysis of prognostic factors of survival in NSCLC patients with brain metastases.METHODS:1. MR imaging and11C-choline PET/CT examinations were performed in30NSCLC patients with brain metastasis or suspected brain metastases.11C-choline uptakes were evaluated by a visual analysis and semi-quantitative analysis using the standardized uptake value (SUV). Analysis the correlation between SUV values and gender, age, pathological type and ECOG score. Calculate the sensitivity and specificity of the various indicators of11C-choline PET/CT imaging in diagnosis of brain metastasis.2.90patients diagnosed with brain metastases underwent treatment and for a long-term follow-up. Analysis the correlations between long-term survival and various clinical factors and biological factors by Cox proportional hazards model, such as gender, age, pathological type, ECOG score, treatment patterns and11C-CHO intake.3. Ninety NSCLC patients with untreated brain metastases were randomly assigned to receive either low-dose temozolomide concomitant with WBRT or WBRT alone. Clincial efficacy and toxicity were compared after a long-time follow up. 4. Pathological examination:Routine HE staining and immunohistochemical examination for biopsy specimens, and observation of biopsy specimens under an optical microscope.5. Statistical analysis:Statistical analysis was performed with the SPSS13.0software. Quantitative results were expressed as mean±SD (X±s), median, and range. The sensitivity, specificity, accuracy, and positive and negative predictive values were calculated. Differences in continuous variables between groups were evaluated using the Student’s t test or the Mann-Whitney rank sum test. Patient groups were compared using the χ2test. The Fisher exact test was used for comparison of frequencies, and Spearman correlation coefficients were calculated to quantify bivariate correlations between data. A receiver operating characteristic (ROC) analysis was performed to compare the diagnostic ability. Cox proportional hazards model was used for multivariate regression analysis. A P-value of<0.05was considered significant.RESULTS:1. Of30NSCLC patients26were confirmed with brain metastases after surgery, magnetic resonance imaging and/or follow-up, with a total of45metastatic lesions. The diagnostic sensitivity of brain metastases with11C-choline is86.7%(39/45), the specificity is85.7%(6/7), th accuracy is86.5%(45/52).2. The means of the four semi-quantitative indicators of SUVmax, L/WM, L/CTX and L/MCU are1.1276、4.1160、9.0769、4.7203, respectively,which have high diagnostic value for brain metastases, with all AUCs>0.9(P<0.01).3. Metastasis diameter has statistical correlation with SUVmax, with a correlation coefficient of0.454(P<0.01). There are no statastical differences of tumor sizes and SUVmax values between different sexes and different age intervals.4. The means of the four semi-quantitative indicators of SUVmax, L/WM, L/CTX and L/MCU in brain metastases with squamous cell carcinom are1.3178,5.1638, 10.3739,5.4774, respectively, and in brain metastases with adenocarcinoma are1.0007、3.4175、8.2119、4.2156, respectively. The identification values of different pathological types with the four semi-quantitative indicators are poor, with AUCs being0.642、0.711、0.626、0.619, respectively. Only the indicator of L/MCU has a certain degree of identification,(P=0.018).5. The response rate was73.3%and77.8%for the WBRT with concurrent low-dose temozolomide arm and WBRT alone arm, respectively(P>0.05). The median survival time was10.0months in the WBRT with concurrent low-dose temozolomide arm and7.7months for WBRT alone arm. There was a survival advantage in the combination arm (P=0.002).6. Headache, rash and alopecia were common in both arms, even more than grade III hair loss was frequent. Hematological and gastrointestinal toxic effects were mild in WBRT alone arm. The most common adverse events in WBRT with concurrent low-dose temozolomide arm were hematologic toxicity and gastrointestinal reactions. More than grade Ⅱ hematological toxicity, gastrointestinal reactions and abnormal liver function were significantly more frequent in the combination arm than those in the WBRT alone arm.7. Cox proportional hazards model analysis in survival data of90patients showed that primary tumor histological type, ECOG score, combined visceral metastasis and simultaneous temozolomide chemotherapy is an independent factor affecting survival (P<0.05), while gender, age and the number of brain metastases has no correlations with survival (P>0.05). The other Cox proportional hazards model analysis in survival data of26patients with brain metastases showed that SUVmax, L/WM and L/MCU affect survival (P<0.05), while L/CTX affect none on that of survival (P>0.05).CONCLUSION:1.11C-CHO PET/CT is very effective in detecting brain metastasis originating from NSCLC, but is limited in identifying pathological types. Pathological type and diameter has statistical correlation with11C-CHO up-take, while sex and age has no statistical correlation with that of11C-CHO.2. WBRT combined with low dose TMZ significantly prolonged survival than WBRT alone for NSCLC patients with brain metastases and was well tolerated. Clinical factors relevant to survival are primary tumor histological type, ECOG score, combined visceral metastasis and simultaneous temozolomide chemotherapy, and biological factors relevant to survival are SUVmax, L/WM and L/MCU. These clinical and biological factors can to some extent nredict prognosis.更多还原