【作者】 李磊；

【导师】 刘吉勇；

【作者基本信息】 山东大学 ， 内科学（专业学位）， 2014， 博士

【摘要】 研究背景随着医疗技术的进步,部分恶性肿瘤的治疗和预后得到了极大改善,但是胰腺癌的治疗和预后仍然处于另一种相对不同的情况。根据最新统计数据,世界范围内胰腺癌死亡率在总癌症死亡率中占第8位,发达国家中占总癌症死亡率的第4位,每年约有266000人死于胰腺癌,并且大部分病人在确诊后的一年内死亡。胰腺位于腹膜后间隙造成胰腺癌早期症状隐匿,直到患者出现明显症状而确诊时疾病多已处于进展期。胰腺癌的预后很大程度上与肿瘤分期有关,早期病人术后5年生存率已经达到20%左右,但是总体的5年生存率仍然不足5%。因此针对胰腺癌危险因素的一级预防在整个疾病的防治中尤为重要。目前认为导致个体罹患胰腺癌的因素大体分为两类：一种是基因因素。关于基因多态性与胰腺癌的关系,人们做了相当深入的研究,也发现了某些基因多态性与胰腺癌的发病风险存在相关性。既往研究显示维生素D(Vitamin D,VD)在体外实验中可以明显抑制胰腺癌细胞的增殖以及口服补充VD的剂量或接受日光照射时间长短与胰腺癌的发病风险相关。但是关于维生素D受体(Vitamin D receptor, VDR)基因多态性与胰腺癌的关系的研究仍非常有限。另一种是环境及后天生理病理因素,如吸烟、饮酒、肥胖、糖尿病以及慢性胰腺炎等。近5年来关于乙型肝炎病毒(Hepatitis B virus, HBV)与胰腺癌相关性的研究逐渐成为胰腺癌流行病学的一个热点,但是不同研究的结果存在争议。我们在进行病例对照研究的基础上汇总既往文献报道结果更新meta分析结果以进一步探讨两者之间的关系。本课题以济南为中心的山东地区胰腺癌病人和健康人群为研究对象,采用适宜的试验方法,对VDR基因rs2228570、rs1544410位点多态性和HBV与胰腺癌发病风险的相关性进行了研究,同时分别探讨VDR基因多态性和HBV与胰腺癌发生部位、疾病分期和病理分化程度的相关性。第一部分维生素D受体基因rs2228570、rs1544410多态性与胰腺癌的相关研究VDR作为一种亲核蛋白受体,广泛分布于体内各组织细胞中。早期研究发现VDR在体内的生物学作用主要是经典的钙磷平衡调节作用,随着近年来研究的不断深入,相关结果显示VDR还具有调节细胞周期、抑制细胞增殖及促进细胞分化和凋亡的作用。基于以上研究,VDR的表达以及VDR基因多态性与癌症的研究成为近十年来的研究热点之一。目前的研究结果显示VDR表达与人类癌细胞株的分化程度有关,对多种组织来源的癌细胞如人前列腺癌细胞、乳腺癌细胞、皮肤癌细胞以及胰腺癌细胞具有调节生长、分化的作用,同时对抑制癌细胞等方面有调节作用。为了进一步探讨VDR基因多态性与胰腺癌的关系,我们设计了本部分研究来分析VDR基因多态性与胰腺癌的发病风险、肿瘤发生部位、疾病分期和癌细胞分化的关系。目的与方法本研究中探讨VDR基因rs2228570、rs1544410位点多态性与胰腺癌的关系。病例组为经病理确诊为胰腺腺癌的患者,对照组为健康查体人群,病例组共258人、对照组385人,病例组与对照组为1：1.5。我们应用限制性片段长度多态性聚合酶链反应(polymerase chain reaction and restriction fragment length polymorphism, PCR-RFLP)技术测定VDR基因rs2228570、rs1544410多态性等位基因以及基因型的频数,并进行单因素以及多因素逻辑回归分析来研究不同等位基因、基因型与胰腺癌发病风险的关系；此外,通过应用列联表分析来检验基因型频率与胰腺癌发生部位、疾病分期和肿瘤分化程度的关系。结果1.rs2228570单核苷酸位点与胰腺癌的关系分别进行单因素回归分析和调整年龄、性别、吸烟史、饮酒史及糖尿病等混杂因素的多因素回归分析,结果显示：rs2228570基因多态性中TT基因型与胰腺癌发病风险显著相关(单因素：OR=3.024,95%CI=1.919～4.764, P=0.005:多因素：AOR=2.978,95%CI=1.844～4.81,P=0.0006):携带至少一个T等位基因的CT+TT基因型胰腺癌患病风险显著增加(单因素：OR=2.424,95%CI=1.718～3.420,P=0.001:多因素：AOR=1.995,95%CI=1.311～3.035,P=0.0013);T等位基因与胰腺癌发病风险显著相关(OR=1.869,95%CI=1.489～2.347,P=0.0001)。将胰腺癌按高分化、中分化、低分化分组,将胰腺癌分化程度与rs2228570多态性基因型频率进行列联表分析,结果显示rs2228570基因多态性与胰腺癌分化程度显著相关(P=0.006,χ2=19.7208)。2. rs1544410基因多态性与胰腺癌的关系分别进行单因素回归分析和调整年龄、性别、吸烟史、饮酒史及糖尿病等混杂因素的多因素回归分析,结果显示：至少含有一个G等位基因的AG+GG基因型与胰腺癌患病风险呈负相关(单因素：OR=0.631,95%CI=0.449～0.887,P=0.008;多因素：AOR=0.648,95%CI=0.453-0.928, P=0.0179)；G等位基因与胰腺癌发病风险呈负相关且有明显统计学意义(OR=0.756,95%CI=0.603～0.947,P=0.015);单独分析AG、GG基因型,未见明显相关性。胰腺癌按TNM分期分为I～Ⅳ期,将胰腺癌分期与rs1544410多态性基因型频率进行列联表分析,结果显示rs1544410基因多态性与胰腺癌分期相关性有明显统计学意义(P=0.0148,χ2＝15.8179)。结论1.VDR基因多态性与个体罹患胰腺癌的风险之间存在显著相关性。rs2228570位点突变纯合子可以导致胰腺癌发病风险增加,突变位点T以及携带至少一个T位点的基因型与胰腺癌发病风险增加有关；rs1544410突变基因位点G则可以降低胰腺癌的发病风险。2.VDR基因多态性与胰腺癌的病理分化、疾病分期等临床特征相关,从而可能影响疾病的预后。第二部分乙型肝炎病毒与胰腺癌相关性的研究以及更新meta分析慢性HBV感染是指感染HBV超过6个月,机体未彻底清除病毒的感染状态。慢性HBV感染可以有明确的急性乙肝病史,也可能由于婴幼儿时期感染HBV,由于当时免疫系统尚未健全无法清除病毒而导致病毒潜伏于机体。近年来关于隐匿性HBV携带状态的研究逐渐增多,临床多以乙肝表面抗原阴性而乙肝核心抗体阳性为特点。我国慢性HBV流行率约为7%,相比较全球感染情况,我国所面临由HBV引起的公共卫生威胁更加严重。近年来的研究表明HBV除定植于肝组织中外,也可以定植于肝外组织比如肾脏、皮肤和胰腺组织中,既往已有数个队列研究和病例对照研究探讨HBV是否可能增加胰腺癌的发病风险,但不同研究所得的结果存在争议。本部分研究中包括胰腺癌病人与健康对照组的病例对照研究；结合病例对照研究结果并汇总以前发表的关于HBV与胰腺癌的研究结果进行的meta分析更新,以进一步探讨HBV与胰腺癌的关系。目的与方法病例对照研究中,病例组为收集的胰腺腺癌病人,对照组为健康查体人群。病例组258人、对照组385人,病例组与对照组为1：1.5。课题中收集了病例组和对照组人群的HBV携带状态、年龄、性别、吸烟史、饮酒史以及糖尿病等临床特征,进行单因素以及多因素逻辑回归分析来研究HBV与胰腺癌发病风险的关系；并结合以上分析结果同时汇总2013年6月以前关于HBV与胰腺癌的研究结果更新meta分析以进一步检验两者之间的相关性；此外,通过应用列联表分析来检验HBV与胰腺癌发生部位、分期和分化程度的关系。结果1.HBV与胰腺癌发病风险的分析回归分析结果显示：乙肝表面抗原阳性组胰腺癌发病风险显著增加(单因素：OR=2.380,95%CI=1.556～3.606,P=0.001;多因素：AOR=3.339,95%CI=2.081～5.358,P=0.001),根据HBV血清学标志物分层研究结果显示HBsAg+/antiHBcAb+组与胰腺癌发病风险显著相关(单因素：OR=3.375,95%CI=1.915～5.950, P=0.001;多因素：AOR=3.864,95%CI=2.057-7.255P=0.0032)；多因素分析显示隐匿性HBV携带组(HBsAg-/antiHBcAb+组)胰腺癌发病风险明显增加(AOR=1.163,95%CI=0.774～1.747,P=0.031)。 meta分析结果显示HBV可以显著增加胰腺癌发病风险(OR=1.392,95%CI=1.167～1.660, P=0.00),亚组分析时队列研究meta亚组分析结果处于临界统计学意义(OR=1.307,95%CI=0.966-1.715, P=0.053)。2.HBV与胰腺癌部位、分期以及分化程度的关系将胰腺癌按高分化、中分化、低分化分组,将胰腺癌分化程度与HBV相关各组进行列联表分析,结果显示HBV与胰腺癌分化程度显著相关(P=0.001,χ2=37.4118)结论1.研究结果提示慢性HBV感染可显著增加胰腺癌的发病风险,隐匿性HBV携带与胰腺癌的发病风险明显相关。以上结论需要进一步多中心、大样本的研究以及不断更新的meta分析进行检验。2.HBV与胰腺癌的病理分化程度相关,从而可能影响胰腺癌疾病的预后。更多还原

【Abstract】 BackgroundsAs medical technology advances, the treatment and prognosis of some forms of cancer have been greatly improved while those of pancreatic cancer (PC) is still in a relatively different situation. According to the latest statistics, PC is the eighth leading cause of cancer-related death in the world and fourth in the developed countries with approximately266000deaths each year. Being one of the most lethal malignant carcinoma in digestive system, PC is characterized by a very poor prognosis due to the asymptomatic nature in early stage and rapid progression thereafter, with most patients dying within12months after diagnosis. Although for the pancreatic cancer patients in early phases, the5-year survival rate after surgery has exceeded to approximately20%, the5-year overall survival rate is only about5%. Therefore the recognition of risk factors and application of primary prevention for pancreatic cancer is of vital importance in the treatment and prognosis of the disease.Risk factors associated with PC can be divided into two main categories. One is the genetic factors. At present, because of the rapid developments in genomics and epidemiology, many studies have proved that some kinds of gene polymorphisms are associated with PC. It has been proved that vitamin D (VD) as a kind of hormone, can significantly inhibit the proliferative activity of many cancer cells included pancreatic cancer cell in vitro and the dietary VD or sunlight exposure may be protective from PC. However, little is known about association about vitamin D receptor (VDR) gene polymorphisms with PC.The other is environmental factors and modifying factors, such as smoking, alcohol, obesity and chronic pancreatitis (CP). In the last5years, some cohort and case-control studies have been conducted to estimate the relationship of hepatitis B virus (HBV) status and PC. The results were controversial.This study is divided into two parts, respectively. In part one, we investigate the relationship of VDR gene rs2228570、rs1544410polymorphisms and PC; in part two, a case control study and an updated meta analysis are performed to discuss the relationship of HBV status and PC.Part1The association of vitamin D receptor gene rs2228570、rs1544410polymorphisms with pancreatic cancerAs one kind of nuclear receptor protein, VDR is widely distributed in different body cells. The first and classical function of VDR is to adjust the calcium-phosphorus balance in blood. As the recently studies, VD and VDR also can regulates cell cycle, inhibit cell proliferation and promote cell differentiation and apoptosis. So many researches have focused on the association of the expression of VDR and VDR gene polymorphisms with different tumors. We conduct the case control study to investigate the relationship between VDR gene polymorphisms and the onset risk, location, TNM classification and pathological differentiation of PC separately. Objective and methodsIn this part, we perform a case control study to value the relationship of VDR gene rs2228570、rs1544410polymorphisms and PC respectively. In this study,258PC patients and385healthy controls were enrolled. The genotypes of rs2228570and rs1544410were assayed using the method of polymerase chain reaction and restriction fragment length polymorphism (PCR-RFLP). Univariate and multivariate logistic regression analysis was done to determine the association of onset risk with gene polymorphisms. Contingency table analysis was done to value the relationship between gene polymorphisms and the location, the TNM classification and pathological differentiation of PC.Result1. rs2228570polymorphisms and PCIn the univariate regression analysis and multivariate logistic regression analysis adjusted by age, gender, history of tobacco, history of alcohol and diabetes, the genotype TT can increase the risk of PC significantly (univariate: OR=3.024,95%CI=1.919-4.764, P=0.005; multivariate:AOR=2.978,95%CI=1.844-4.81, P=0.0006); the genotype with at least one T gene loci can increase the risk of PC significantly (univariate:OR=2.424,95%CI-1.718-3.420, P=0.001; multivariate:AOR=1.995,95%CI=1.311-3.035, P=0.0013); the frequencies of T allele in PC group was significantly higher than in controls (OR=1.869,95%CI=1.489-2.347, P=0.0001). In the contingency table analysis, the rs2228570polymorphisms were correlated with pathological differentiation of PC significantly.2. rs1544410polymorphisms and PCIn the univariate regression analysis and multivariate logistic regression analysis adjusted by age, gender, history of tobacco, history of alcohol and diabetes, the genotype with at least one G gene loci could decrease the risk of PC significantly (univariate:OR=0.631,95%CI=0.449-0.887, P=0.008; multivariate: AOR=0.648,95%CI=0.453-0.928, P=0.0179); the frequencies of G allele can decrease the risk of PC significantly (OR=0.756,95%CI=0.603-0.947, P=0.015). In the contingency table analysis, the rs1544410polymorphisms were correlated with TNM classification of PC significantly.Conclusion1.In this study, the VDR gene polymorphisms were significantly correlated to with the onset risk of PC. In rs2228570, the T loci and genotype with T allele could increase the risk of PC; in rs1544410, the G loci and genotypes AG+GG could decrease the onset risk of PC significantly.2.The VDR gene polymorphisms were correlated with pathological differentiation and TNM classification of PC significantly. So these polymorphisms might affect the prognosis of this disease.Part2The association of hepatitis B virus status with pancreatic cancer including a case control study and an updated meta analysis.HBV is still responsible for heavy disease burdens in China. Chronic HBV status refers to the infection of HBV for more than6months, this infection status could attribute to a history of acute HBV infection or infant period HBV infection in which the hepatitis B virus are not completely cleaned up due to the imperfect development of the immune system. In2006, according to the Ministry of Health of China, the national prevalence of HBV was about7%. And today, more and more researches were focused on the occult HBV status which is significantly associated with simply presence of anti-HBc antibody. Several studies found that HBV could replicate in human pancreatic tissue and that patients with HBV infection have impairments of pancreatic function. Some cohort and case-control studies have been conducted to estimate the relationship between HBV status and PC, providing somewhat conflicting results. In this part, we conducted a case control study and update the meta analysis to investigate the relationship of HBVstatus and PC. Objective and methodsIn this part, we perform a case control study and a meta analysis to further value the relationship of HBV status and PC. In the case control study,258PC patients and385healthy controls were enrolled. Nine studies were included to perform the meta analysis together with the result of our case control study. The characteristics of the patients and controls were collected such as age, gender, history of tobacco and alcohol exposure and diabetes. Univariate and multivariate logistic regression analysis was done to determine the association of PC onset risk and HBV; contingency table analysis was done to value the relationship between HBV and the location, the TNM classification and the pathological differentiation of PC. The adjusted RRs, HRs or ORs were collected to compute a summary OR and95%CI. HRs, RRs were directly considered as ORs. Heterogeneity across studies was tested using the Q and I2statistic. Stratified meta analysis was conduct according to region and study design.ResultIn the univariate regression analysis and multivariate logistic regression analysis adjusted by age, gender, history of tobacco, history of alcohol and diabetes, the chronic HBV infection can increase the risk of PC significantly (univariate:OR=2.380,95%CI=1.556-3.606, P=0.001; multivariate: AOR=3.339,95%CI=2.081-5.358, P=0.001), in the the stratified analysis according to HBV serological markers, the group HBsAg+/anti-HBcAb+can increase the risk of PC significantly; the occult HBV carrier status also can increase the risk of PC significantly, the AOR was1.163(95%CI=0.774-1.747, P^O.031). In the meta analysis, the overall OR was1.392(95%CI=4.167-1.660, P=0.00) assembly to the result of submeta analysis according to region. Only a borderline significant association was observed when combining all cohort studies (OR-1.307,95%CI=0.966-1.715, P=0.053). Given cohort study was more powerful to detect a causal relationship than case-control one, this result should be interpreted with caution. In the contingency table analysis, the HBV status was correlated with pathological differentiation of PC significantly. Conclusion1.The chronic HBV infection and occult HBV carrier might increase the onset risk of PC significantly, but further multi-center large sample clinical studies especially cohort studies are required to investigate the relationship.2. The HBV status was correlated with pathological differentiation of PC significantly. So hepatitis B virus might affect the prognosis of the disease.更多还原