【作者】 居旻杰；

【导师】 汤钊猷；

【作者基本信息】 复旦大学 ， 外科学， 2009， 博士

【摘要】 原发性肝癌（主要为肝细胞癌,HCC,简称肝癌）在全球恶性肿瘤死因中居第3位,在我国更高居第2位。尽管肝癌临床和基础研究均取得了长足的进步,但其总预后并无显著改善。即使根治性切除5年复发转移率仍高达60-70%;局部治疗则更高。为此,转移复发已成为进一步提高疗效的瓶颈。传统的研究注重癌细胞自身,对宿主及局部微环境研究较少。近年发现宿主微环境对肿瘤演进起着重要作用。目前已认识到癌转移是肿瘤细胞和宿主微环境之间相互作用的过程,后者能影响肿瘤的生物学特性并促进或抑制肿瘤细胞在靶器官的定居。随着芯片技术等高通量研究方法的涌现,大量来源于微环境、参与肿瘤器官特异性转移的分子、基因表达谱及信号通路等被发现。宿主微环境中,复发转移靶器官的微环境是影响复发转移是否发生的决定性因素之一,癌细胞只有在适合的靶器官微环境中才能生长,形成了肿瘤转移复发的“器官特异性”,百年前Paget以“种子与土壤”理论对此做了精辟的总结。在宿主因素中,许多是与炎症/免疫相关的因素,靶器官的炎症/免疫反应对患者预后有重要影响,但有双向调节的特性。如能增强不利于肿瘤转移的靶器官炎症状态或削弱有利于转移形成的状态,当可降低转移的发生。但前提是需判断靶器官处于何种炎症/免疫状态,以及何种状态可促转移或抑转移,否则干预措施将是盲目的。有乙型或丙型肝炎病毒感染背景的肝癌,是研究炎症与癌症关系的重要对象。肝脏有独特的免疫系统并参与局部及整体水平炎症、免疫调节。肝内不仅定居着大量的淋巴细胞,还有许多炎症相关细胞,这其中肝星状细胞在肝脏炎症反应向纤维化和癌变转化的整个过程中都起着重要做用。肝内的炎症及炎症相关细胞在生理状态下通过与循环中的淋巴细胞的相互作用,使肝内的炎症/免疫反应处于动态平衡中。而在病毒感染等病理情况下,它们会引起肝脏炎症反应的慢性迁延以及由此产生的纤维化甚至癌变。我所前期研究发现肿瘤内的炎症/免疫反应状态能影响肝癌细胞侵袭转移潜能,从而影响预后。肝癌根治性切除后余肝是术后转移复发的主要靶器官,癌旁肝组织很好的代表了残余肝,为此癌旁肝炎症状态应具有预后价值。已有研究提示癌旁肝组织的免疫状态与肝癌门静脉播散相关。但目前尚无针对癌旁炎症免疫状态的系统研究,也没有简便可靠的判断指标,从而无法给予治疗以准确的监测与指导。综上所述,针对肝脏炎症的治疗,当有助抑制肿瘤复发转移。根据“种子与土壤”的观点,在针对肝脏炎症的同时,还应针对肿瘤细胞,才能发挥更大作用。干扰素-α（interferon-α,IFN-α）既有治疗乙型和丙型肝炎病毒的作用,减轻炎症反应;又有抑制肿瘤细胞增殖、侵袭转移的作用;并已被证实对肝癌术后患者有预防或延迟复发的作用。但干扰素治疗又有明显的不足:一方面患者常需大剂量使用,易引起明显的副作用,使患者的依从性降低;另一方面,IFN-α在诱导细胞凋亡的同时,也能促进细胞存活;对于肾癌、膀胱癌、非小细胞肺癌等恶性肿瘤等更存在针对IFN-α的耐药性,这种耐药性可能与其诱导环氧化酶2（COX-2）表达上调有关。通过联合使用中药姜黄素或COX-2抑制剂（如塞来考昔）等,能逆转IFN-α上调的COX-2表达,提高其疗效。IFN-α对肝癌也存在耐药性,仅对部分术后肝癌患者有效。因此有必要寻找减少干扰素用量,同时能提高其疗效的方法。COX-2抑制剂塞来考昔等虽可选用,但有心血管反应等副作用,因此其临床使用受到限制。中药制剂通常作用温和、副作用较小,其中氧化苦参碱（苦参素,oxymatrine）是临床常用的清热解毒药,已被用于抗乙肝的治疗,能抑制病毒复制,效果与IFN-α相似;又有调节免疫、抗肝纤维化等作用、抗肿瘤等作用。体外实验发现其能抑制肝痛细胞的增殖、促进凋亡。为此氧化苦参碱也具有调节炎症反应和抗肿瘤细胞的双重作用,故小剂量IFN-α联合氧化苦参碱可能有较好的临床效果。COX-2表达的上调可能是进一步提高干扰素对肝癌疗效的瓶颈,但尚无相关研究。研究提示,在慢性肝病患者中COX-2可调节炎症局部T淋巴细胞和巨噬细胞的浸润;促进慢性肝炎向肝纤维化/肝硬化的转变;COX-2更是慢性肝炎反应向肝癌转变的重要参与者和调控者,癌变过程中HBV基因组向宿主肝细胞内的整合需要COX-2的参与;选择性的抑制COX-2则能抑制肿瘤的进展并有一定的预防疗效。但COX-2对于肿瘤和炎症反应来说是一把双刃剑,既能促进肿瘤生长、加剧慢性炎症反应,又能抑制肿瘤生长,抑制慢性炎症反应。目前尚无癌内COX-2与肝癌预后的研究。所以必须首先明确COX-2与肝癌侵袭转移及预后的关系,才能以此为基础研究其在肝癌干扰素治疗中的作用。本文目的是:阐明肝脏炎症与肝癌转移的关系,探索小剂量IFN-α为基础的干预措施;明确COX-2与肝癌转移、预后以及小剂量IFN-α疗效的关系,以提高IFN-α疗效。关于判断肝脏炎症反应状态,本文从两方面入手:一是以临床易行的外周血中与肝脏炎症密切相关的γ-谷氨酰转肽酶（GGT）和谷氨酸氨基转移酶（ALT）为对象;二是以有助说明机理的癌旁肝组织中重要的炎症相关细胞为对象。方法上用组织微阵列免疫组化技术和荧光定量PCR技术,原位、在体探索活化肝星状细胞、肥大细胞、Foxp3+调节性T细胞和巨噬细胞等的分布、数目及相关基因表达与肿瘤转移的关系。关于干预,本文进行了以小剂量IFN-α为基础的干预研究。第一部分外周血γ-谷氨酰转肽酶、谷氨酸氨基转移酶与肝癌患者术后预后目的:寻找临床可用、简便易行的反映肝脏炎症和肝癌预后的预测指标。方法:随机选取了2002年1月—2006年12月,219例在复旦大学附属中山医院行根治性手术切除的肝细胞癌患者。除肝癌外,所有患者术前无其他影响GGT和ALT水平的疾病,肝功能Child-Pugh分级均为A级,均在术前两天采集外周静脉血检测肝功能。GGT或ALT分别分为高低两组（GGT的分界值为60IU/L,ALT的分界值为80IU/L）,比较高低GGT或ALT组患者的术后生存时间（OS）和至复发时间（TTR）的差异。还计算了GGT与ALT的比值（GGT/ALT）,并将其按最佳分界值（恰为其平均值）分为高低两组,同样比较了两组间OS和TTR的差异。结果:外周血GGT与ALT之间具有显著的线性正相关（r=0.411,P＜0.001）,但当将GGT和ALT都分为高低两组后,多数患者的GGT处于高水平（n=110）而ALT处于低水平（n=185）。GGT越高,患者肿瘤越大,BCLC分期越晚,越有可能发生血管侵犯。而ALT则与肿瘤临床病理特征无明显相关性。而GGT/ALT表现出与肿瘤病理特征更为紧密的联系,其高低与肿瘤大小、有无包膜、有无血管侵犯、BCLC分期都显著相关。生存分析发现,GGT仅与OS具有独立的相关性（P=0.002）,ALT与TTR和OS都无相关性;而GGT/ALT与TTR和OS均独立负相关（P值分别为0.002和小于0.001）。重要的是GGT/ALT越高者越容易发生术后早期肝内复发（P＜0.05）和复发相关死亡（P＜0.01）,而与术后晚期复发及非复发相关死亡的发生无明显相关性。将患者按术前血清甲胎蛋白（AFP）高低分为两组后,GGT/ALT在这两组病人中,仍具有一定的预后预测能力,通过受试者工作曲线（ROC）分析,发现GGT/ALT预测预后的能力高于AFP。结论:以外周血GGT和ALT联合分析（GGT/ALT）为指标判断肝脏炎症反应状态能较好地预测肝癌患者的术后预后。鉴于GGT和GGT/ALT与肿瘤临床病理特征紧密相关,GGT/ALT也与早期复发相关,为此这些指标不仅可反映肝脏炎症状态,也能反映患者肿瘤负荷（如大小,有无播散等）。提示肝脏炎症反应与HCC的复发转移相关,以GGT或GGT/ALT为指标,有助筛选术后复发高危患者并给予干预,有助监测治疗后肿瘤负荷和肝脏炎症的动态变化,并指导治疗的调整。第二部分癌周炎症（相关）细胞与肝癌患者术后预后目的:研究有助反映肝脏炎症和肝癌预后的癌旁肝组织中重要相关细胞,作为术后更确切的预测指标。方法:随机选取2002年2月—2005年11月,130例（队列A）和2002年1月—2006年12月,207例（队列B）两组在复旦大学附属中山医院行根治性手术切除的肝细胞癌患者。队列A:首先通过组织微阵列,利用免疫组化染色和细胞数目绝对定量,原位、在体地评价了癌旁微环境中αSMA+活化的肝星状细胞（aHSC）的数目、分布、功能状态与术后肿瘤复发转移及肿瘤侵袭能力指标之间的联系;然后利用荧光定量RT-PCR,检测这130例患者癌旁冰冻标本中主要由aHSC表达的功能基因包括粘蛋白-C（TNC）、骨连接素（SPARC）和纤维母细胞活化蛋白（FAP）的表达量,以及上述功能基因与肝癌患者术后预后的关系。队列B:利用组织芯片、免疫组织化学技术,绝对计数癌旁tryptase+肥大细胞（MC）的数目、分布、功能状态与术后肿瘤复发转移及肿瘤侵袭能力指标之间的联系。在队列A和队列B中,通过免疫组织化学和细胞绝对计数方法,研究了癌旁Foxp3+调节性T细胞（Treg）、CD68+巨噬细胞（MΦ）等炎症免疫相关细胞的数目、分布和功能状态与肝癌患者预后的关系,以及aHSC、MC与Treg、MΦ等在数量和功能上的关系。结果:（1）癌旁Treg和MΦ的数量与aHSC数量成正相关。（2）aHSC功能基因FAP、SPARC和TNC的表达水平之间存在显著的正相关,而且它们的表达水平都与aHSC数量正相关（r分别为0.903、0.634和0.887,P＜0.001）;与正常肝组织相比,癌旁肝组织的上述功能基因的表达量更高。（3）癌旁aHSC数量（r=0.176,P=0.04）与功能基因（FAP）（r=0.204,P=0.02）表达水平与外周血GGT/ALT水平正相关;癌旁aHSC数量、FAP和SPRAC的表达量都与外周血GGT含量正相关（r分别为0.222、0.287和0.174,P分别为0.011、0.001和0.047）。（4）癌旁aHSC在数量和功能基因的表达量上,都与肝癌恶性临床病理特征正相关。（5）癌旁aHSC的数量与功能基因的表达高低都与术后生存和复发相关,即数量高或功能基因表达高者术后生存时间短,更易复发;同样癌旁Treg和MΦ也是患者预后的独立预测指标。（6）癌旁aHSC的数量与功能基因表达量、Treg及MΦ数量联合后,具有更好的预测能力。（7）癌旁aHSC的数量与功能基因表达高者更易发生早期复发,而对晚期复发无显著影响。（8）癌旁MC与癌旁Treg在数量上存在显著的正相关性（r=0.353,P＜0.001）。（9）MC数量与外周血GGT/ALT含量正相关（r=0.146,P=0.036）。（10）MC单独使用时,仅能独立地预测术后复发,而与总生存时间无关;在联合Treg分析后,其预后价值显著提高,癌旁MC和Treg同时高的患者预后显著差于MC和Treg都低者。而且MC数量的高低与早期复发相关,而与晚期复发无关。（11）MC的数量与肿瘤临床病理特征相关,MC越多,肿瘤越大,越有可能没有完整的包膜。结论:（1）肝癌癌旁微环境aHSC在细胞水平和基因水平均可作为患者术后复发转移和总生存时间的预测指标。癌旁aHSC能通过表达功能基因如FAP、SPARC等,加剧肝脏炎症反应,诱导Treg和MΦ在癌旁肝组织聚集,形成有利于播散肿瘤细胞定居的炎症免疫环境,增加播散肿瘤细胞侵袭转移活性,更易形成早期肝内复发。当aHSC的数量与其功能基因或其他炎症/免疫细胞联合后,具有更好的预测价值,提示活化的aHSC主要通过调节肝脏炎症/免疫反应来加速肝癌术后复发转移。以aHSC为靶点的抗炎/免疫调节治疗具有潜在预防复发转移、改善预后的价值。（2）癌旁MC也可通过调节炎症免疫反应,形成有利于肝癌复发转移的土壤,促进播散肿瘤细胞的侵袭转移。（3）癌旁aHSC在细胞数量和功能基因的表达水平上与外周血GGT和/或GGT/ALT水平成正相关;癌旁MC数量与外周血GGT/ALT水平正相关。提示以癌旁炎症相关细胞（活化的肝星状细胞、肥大细胞等）来判断肝癌患者的肝脏炎症反应,和以外周血肝酶（GGT和GGT/ALT）判断患者肝脏炎症状态,其预测结论一致,从而反证简便易行的GGT和GGT/ALT的预测价值。我们认为,通过两方面的综合分析更能客观地反映肝脏炎症反应的状态。第三部分环氧合酶-2与肝癌侵袭转移潜能和预后的关系目的:验证环氧合酶-2（cyclooxygenase-2,COX-2）与肝癌细胞侵袭转移潜能和肝癌预后的关系,为进一步研究小剂量IFN-α的疗效,及其疗效与COX-2的关系打下基础。方法:以不同侵袭转移潜能的七种肝癌细胞株和随机选取的根治性肝癌切除术后患者为研究对象（排除晚期、肝外复发转移）,通过Western-blot和荧光定量PCR技术,分别测定肿瘤细胞和肿瘤组织内COX-2的表达,分析COX-2表达与肿瘤细胞侵袭转移潜能和肿瘤患者预后的关系。结果:（1）随着肿瘤细胞侵袭转移潜能的降低,其COX-2表达量同步减少。（2）肝癌组织COX-2 mRNA表达高低与主要临床病理特征无显著联系。（3）肿瘤内COX-2表达越高,患者复发和死亡的危险越高,P值分别为0.025和小于0.001,危险度分别为1.916（1.086-3.380）和3.522（1.985-6.248）。结论:COX-2高表达促进肝癌侵袭转移潜能,加速早期肝内复发,使预后恶化。如干扰素能抑制肝癌表达COX-2,则能抑制其侵袭转移,降低复发;反之,则增强肝癌侵袭转移,造成不良预后。第四部分小剂量干扰素单用或联合氧化苦参碱对MHCC97-H细胞侵袭转移潜能的影响及可能的机制目的:根据“种子与土壤”的观点,理想的干预应既可调节肝脏炎症又能抑制肿瘤。IFN-α虽被证实可预防术后复发转移并有抗炎免疫调节作用,但较大剂量使用其副作用明显。本研究拟通过体外试验,探索小剂量干扰素作为基础干预的可能性,以及提高其疗效的途径。方法:研究小剂量IFN-α对高转移潜能人肝癌细胞MHCC97-H的作用,并研究氧化苦参碱与之合用后是否有更强的抑瘤作用。根据文献,干扰素小剂量标准为1×10³U/ml,我们研究该剂量干扰素对MHCC97-H细胞体外增殖、凋亡、侵袭转移能力的影响。同时,研究了小剂量干扰素与氧化苦参碱联合使用后对MHCC97-H的作用,以明确氧化苦参碱与干扰素之间是否具有协同的抗肿瘤作用及可能机制。结果:（1）小剂量干扰素体外对MHCC97-H细胞的增殖抑制能力不显著,且不随时间增加;不能显著的诱导MHCC97-H细胞凋亡,且不随时间延长而增加;不能明显抑制MHCC97-H细胞体外侵袭转移潜能。（2）氧化苦参碱对MHCC97-H细胞具有一定的增殖抑制能力,且随浓度增加而增加,其IC₅₀=44.5mg/ml,为了避免选择较高增殖抑制能力的氧化苦参碱浓度而掩盖了后面判断联合用药的效果,我们选择具有较小增殖抑制能力的氧化苦参碱浓度=10mg/ml作为进一步实验的浓度。（3）联合用药后,对MHCC97-H细胞的增殖抑制能力显著增强,且随作用时间延长而增加;并能显著地诱导MHCC97-H细胞凋亡,同样随时间延长而增强;亦能有效地降低MHCC97-H细胞的侵袭转移能力。（4）机制:小剂量干扰素具有上调肿瘤细胞COX-2表达的作用,而氧化苦参碱则能抑制COX-2表达,两者联合用药后,小剂量干扰素所上调的COX-2表达亦被有效抑制。结论:小剂量干扰素体外对MHCC97-H细胞没有明显抑制增殖、促进凋亡和降低侵袭力的作用,这与其上调COX-2表达有关;而合用氧化苦参碱后,其作用增强,且伴COX-2表达的显著下调。说明氧化苦参碱通过逆转由小剂量干扰素上调的COX-2表达而产生协同抗肿瘤作用。提示两者的联合可能降低干扰素的耐药性,减少干扰素使用剂量。第五部分体内小剂量干扰素单用或联合氧化苦参碱对肝脏炎症状态的影响及对原位MHCC97-H种植瘤的作用目的:通过高转移人肝癌裸鼠模型进一步在体内验证小剂量干扰素合用氧化苦参碱对肿瘤和肝脏炎症的作用。方法:采用高转移人肝癌细胞系MHCC97-H裸鼠原位移植模型。干扰素用量为1×10⁶U/ml,0.1ml/d皮下注射;氧化苦参碱为60mg/kg·d,腹腔注射。裸鼠分4组:空白组（n=12）皮下及腹腔注射生理盐水62天;干扰素组（n=8）腹腔注射生理盐水,皮下干扰素,连续62天;氧化苦参碱组（n=8）皮下注射生理盐水,腹腔氧化苦参碱,连续62天;联合治疗组（n=6）皮下干扰素加腹腔氧化苦参碱62天。动物若在治疗期间死亡（对照组3只,干扰素组2只,氧化苦参碱组2只,联合治疗组无死亡）视为观察终点,用于比较生存期;动物存活至治疗结束者,处死前取外周血检验肝功能,处死后取组织标本行相应实验。另设同样四组,每组12只裸鼠,用于观察生存期,故最后观察生存期的各组裸鼠数量分别为:空白组15只,干扰素组14只,氧化苦参碱组14,联合治疗组12只,最长生存时间为97天。结果:（1）小剂量干扰素组裸鼠体重有下降趋势,但差别不显著;氧化苦参碱则有增加裸鼠体重的趋势;两者合用后,裸鼠体重明显增加。（2）就种植肿瘤的体积而言【与对照组比较（3.056±0.649cm³）】,小剂量干扰素组肿瘤最大（6.410±1.854cm³,P=0.055）,而联合用药组肿瘤最小（0.382±0.269cm³,P=0.007）。（3）与对照组相比,仅联合治疗组的肺转移率显著下降（P=0.011）;但计算肺转移灶个数后,发现氧化苦参碱组（P=0.02）和联合治疗组（P=0.008）的肺转移个数均显著少于对照组,而小剂量干扰素组则显著增加（P=0.018）。（4）联合治疗组的生存时间显著长于其他各组,而小剂量干扰素组未能提高生存时间,氧化苦参碱组与小剂量干扰素组相比无显著差异。（5）小剂量干扰素作用后,外周血GGT和GGT/ALT均显著提高,而联合治疗组则显著降低。（6）各组外周血GGT及GGT/ALT均与其肿瘤大小正相关。（7）癌旁和癌内活化的HSC在小剂量干扰素组数量最多,联合治疗组数量最少。（8）癌内COX-2的表达在对照组和小剂量干扰素组较高,而在氧化苦参碱组和联合用药组则明显降低。结论:（1）小剂量干扰素不能抑制移植瘤生长,反增加肺转移灶数目,未能延长生存期。合用氧化苦参碱能显著抑制移植瘤的生长和肺转移,延长生存期。（2）可能的机制:A、抑制肝脏炎症反应,联合用药后外周血GGT、GGT/ALT显著降低,而小剂量干扰素则使之提高;且GGT、GGT/ALT在各组均与肿瘤大小正相关,说明GGT/ALT或GGT能反映体内肿瘤负荷。联合治疗组癌旁及癌内的aHSC数量减少,提示肝脏炎症反应程度降低。B、抑制肿瘤侵袭转移能力:联合用药后肺转移率和转移灶个数都显著降低。可能因氧化苦参碱能逆转小剂量干扰素诱导的肿瘤细胞COX-2表达,从而抑制肿瘤侵袭转移能力,降低肺转移率。而COX-2通路又是炎症反应中的重要通路之一,故两药联合在抑制肝脏炎症反应方面同样具有协同作用。为此,体内研究证明,联合用药通过苦参碱逆转由小剂量干扰素诱导的COX-2,从而更有效地抑制炎症反应并抑制肿瘤。结论1.肝脏炎症状态对于HCC患者具有重要的预后价值。通过研究外周血GGT和GGT/ALT,以及癌旁炎症相关细胞的数量和功能状态,均能反映肝脏的炎症反应状态,并预测患者的预后,两者结论一致。前者简便易行,还可连续监测治疗并指导其调整。后者有助说明机理,并为治疗提供潜在的靶点。2.小剂量干扰素的抗瘤和抑制肝脏炎症的作用不明显。小剂量干扰素与氧化苦参碱合用则显著抑制肝癌的生长和侵袭转移,与氧化苦参碱能逆转由小剂量干扰素上调的COX-2表达有关。3.联合用药更能显著地降低外周血GGT,GGT/ATL水平和癌旁活化HSC数量,从而改善肝脏的炎症状态,形成不利肿瘤生长之环境,抑制转移复发形成。创新点1.首次发现和报道了肝癌患者外周血GGT/ALT与术后复发转移的密切关系,该比例可以作为连接肝脏炎症反应与肿瘤负荷的桥梁,作为筛选高危患者、及时进行术后干预的指标,也可作为术后炎症免疫调节治疗的监测指标。（Journal ofGastroenterology 2009）2.首次发现和报道了肝癌癌旁微环境中活化肝星状细胞（Am J Clin Pathol2009）和肥大细胞（Cancer Science 2009）与术后复发转移和生存时间的密切联系,同样有助于筛选高危复发患者,并提供了术后干预治疗的靶点。3.通过体内外实验,首次发现氧化苦参碱与小剂量干扰素合用,可抑制肝脏炎症反应并抑制肝癌的生长与侵袭转移,其协同作用与氧化苦参碱能逆转小剂量干扰素所诱导的COX-2表达有关。而小剂量干扰素既不能抑制肝癌生长、侵袭和转移,又不能调节肝脏炎症反应。潜在应用价值1.外周血GGT/ALT比例及癌旁以活化肝星状细胞和肥大细胞为代表的炎症（相关）细胞都能较好地反映肝癌患者肝脏炎症状态,并与术后复发转移和生存密切相关,可用于临床筛选高危复发患者、指导和监测术后治疗。2.不能有效的改善肝脏炎症状态是小剂量干扰素抗肿瘤、抑制复发转移疗效局限的重要原因之一。小剂量干扰素联合氧化苦参碱可作为肝癌术后又一辅助治疗的方案。更多还原

【Abstract】 Primary liver cancer is one of the most common malignant tumors,and is currently the third and second leading cause of tumor-related death worldwide and in China,respectively.Hepatocellular carcinoma（HCC） undoubtedly contributes to the majority primary liver cancer cases.Despite tremendous achievements being carried out in HCC clinically and basically during past decades,the overall prognosis of HCC remains dismal with a 5-year survival at about 5%.Even after curative resections,the 5-year recurrence/metastasis rates remain as high as 60%-70%.Therefore,recurrence or metastasis has emerged as the main obstacle to get better curative effects.The vast majority of previous studies focused solely on malignant cells themselves,regarding tumors just as masses of autonomous cells and aiming at identifying the molecular and genetic changes associated with this malignant transformation.However,the potential of a tumour cell to metastasize depends on its interactions with the homeostatic factors that promote tumour cell growth,survival, angiogenesis,invasion and arrest in distant organs.Given the "site-specific metastases" characteristic of tumor metastases,Stephen Paget had developed the theory of "seed and soil" in 1898,and recently,with the advent of high throughput analyses like genomic,protemics and tissue microarray,numerous growth factors, chemokines,guidance molecules,signaling pathways and,more importantly,new genes in host microenvironment are discovered and provide a new identity for organ-specific metastasis.Inflammatory/immune related elements contribute a great deal to the establishment of host microenvironment and modulate the progression of cancers.For instance,chronic inflammation is capable of generating a potentially "vicious self-sustaining loop（s）" which are resulted in the pro-cancer microenvironment favorable for survival of tumor cells and their growth.Therefore. the inflammatory responses of the target organs may be also of great significance regarding to the prognosis of cancer patients.However,inflammation can either promote or inhibit cancer metastasis/ recurrence.If we can modulate this double-edged sword of inflamrnation rigorously in order to impress the favorablc host inflammatory microenvironment or irritate unfavorable one for tumor metastasis, tumor recurrence may be under control.To achieve this goal,it is indispensable for us to get a panorama about the status of host inflammatory response,and to understand the accurate relation between one certain inflammation status and tumor recurrence/metastasis;otherwise,any intervention targeted at inflammation response will be aimless.The liver has its own unique immune system,acting as a key immune regulator locally and systemically.In addition,considering the background of hepatitis B and/ or C virus infection,the involvement of inflammation in HCC is of great significance. Among the abundant resident lymphocytes and inflammatory-related cells,hepatic stellate cell（HSC） is significantly involved in the progression of chronic hepatitis to hepatic fibrosis or even liver carcinogenesis.All these inflammatory or inflammatory-related cells are critical for the maintenance of immune homeostasis via their interaction with the circulating lymphocytes.However,in the context of infections,these cells will protract the inflammation reaction and inevitably result in fibrosis or carcinogenesis.For instance,our previous study has proved that the intratumoral inflammation/ immune status could influence the invasion and metastastic capability of hepatoma cells and was related to the prognosis of HCC. More importantly,the recurrence/metastasis of HCC after curative resections mostly occur within the residual liver tissues;hence liver itself is the major target organ for HCC recurrence/metastasis.It is reasonable for us to suggest that the inflammation response in the peritumoral liver tissue is likewise indispensable for HCC recurrence/ metastasis.Unfortunately,no systemic evaluations of the peritumoral inflammation response and its relevance to HCC recurrence have been carried out yet.As a result. there is no available information for an accurate monitoring and guiding purpose for postoperative therapies.Hepatic inflammation-oriented interventions should be helpful in suppressing HCC recurrence/metastasis.According to the "seed and soil" theory,therapies should be targeted not only against the cancer cells themselves,but also against the homeostatic inflammatory factors,lnterferon-α（IFN-α） can both ease the inflammation responses of HCV and/or HBV infection,and inhibit the proliferation, invasion and metastasis of tumor cells in vitro.Clinically.IFN-αhas been proved to be able to postpone the recurrence of HCC after curative resection.Therefore,IFN-αfits well with above-mentioned criteria.However,these contributions are based upon the long-term application with a high-dose of IFN-αwhich is near the toxic dose and related to visible by-effects.Moreover,the resistances to IFN-αin renal cancer, gladder carcinoma and so on have been detected.Cycloxygenase-2（COX-2） has been proved to be the key in IFN-αresistance,and agents such as curcumine and celecoxib can suppress COX-2 expression and improve the effect of IFN-α.With respect to HCC,there was also resistance to IFN-α.So,it is urgently demanded to find a way to decrease the dose of IFN-αmeanwhile improve its effect.Although chemosynthesis agents such as celecoxib are available candidates,their clinical utility is restricted by their side effects including cardiovascular complaints.The traditional Chinese medicine,however,is another choice,for its gentle effects with little side effects. Oxymatrine is now widely employed in the treatment for HBV with an effect similar to that of IFN-α.Besides,oxymatrine can modulate immune activity,and inhibit hepatic fibrosis and tumor progression.In vitro,oxymatrine were able to inhibit proliferation and induce apoptosis of liver cancer cells.Hence,oxymatrine,similar to IFN-α,can both modulate hepatic inflammation and repress tumor cells.We suggest that IFN-αin combination with oxymatrine may display better anticancer effects.COX-2 is heavily involved in IFN-αresistance.However,its role in HCC IFN-αtreatment is still unknown.In chronic liver diseases,COX-2 is unregulated to induce local infiltration of T lymphocytes and macrophages.Additionally,COX-2 is actively involved in HCC initiation via facilitating the integrating of HBx gene into host hepatocytes.Using COX-2 inhibitors can usually block HCC progression and improve the prognosis.However,the roles of COX-2 in liver inflammation,fibrosis and carcinogenesis are paradoxical.Therefore,a better understand of the role of COX-2 in HCC IFN-αtreatments should been based upon a thorough understanding of the relationship between COX-2 and HCC invasion,metastasis and prognosis.The aim of this study is（a） illuminate the correlation between hepatic inflammation and HCC metastasis;（b） estimate the effect of low-dose IFN-αin HCC treatments;（c） explore the role of COX-2 in IFN-αtreatments based upon identification the involvement of COX-2 in HCC metastasis and prognosis.The hepatic inflammation status is evaluated both with periphery parameters includingγ-glutamyltransferase（GGT）,alanine aminotransferase（ALT） and their ratio （GGT/ALT）,and with peritumoral inflammation（related） cells including activated hepatic stellate cell（aHSC） and mast cell（MC） by tissue microarray（TMA） and quantitive RT-PCR methods.With respect to the interventions,low-dose IFN-αis used both in vitro and in vivo to determine its influence over HCC.In addition. oxymatrine is utilized in combination with low-dose IFN-αto exam whether this combination can improve IFN-αeffect or not,and to evaluate its effect over tumor cells and hepatic inflammation.PartⅠ.The relevance of peripheral GGT and ALT to HCC prognosisPurpose:To investigate the involvement of hepatic inflammation in HCC reucrrecne/metatstasis by evaluating peripheral GGT,ALT and GGT/ALT.Methods:In this study we enrolled one independent and random cohort of 219 HCC patients underwent curative resection in Zhongshan Hospital,Fudan University from Jan 2002 to Dec 2006.All the patients with other diseases which can influence the level of GGT and ALT were excluded.The live reserve functions of all patients are in Child-Pugh A stage.The serum samples were taken two days before operations and the normal rang of GGT and ALT were 11-50 IU/L and＜75 IU/L,respectively. According to the documents,60 and 80 IU/L were used as the cutoff points for GGT and ALT,respectively.The overall survival（OS） and time to recurrence（TTR） were compared between high and low subgroups of GGT and ALT.Moreover,we calculated the ratio of GGT to ALT（GGT/ALT）,and divided GGT/ALT into high and low subgroups by a optimal cutoff point which was created by X-tile software. The prognostic value of GGT/ALT was examined as well.Results:GGT and ALT were related to each other in a linear fashion（r=0.411,P＜0.001）.However.as dichotomized variables,GGT and ALT were distributed unequally（P＜0.001）.Most patients had high levels of GGT（n=110） while low levels of ALT（n=185）.ALT displayed no relation to recurrence or survival,while GGT was independently associated with survival（P=0.002）.The GGT/ALT ratio could predict survival precisely either in a continuous or dichotomized fashion（P＜0.001 and=0.001.respectively）,and also related to recurrence when dichotomized（P =0.002）.Additionally,high GGT/ALT ratio was associated with high early-recurrence rates,more recurrence-related death and various aggressive tumor characteristics such as lager tumor size,vascular invasion,poor encapsulation and advanced BCLC stage.In further stratified analyses,this ratio could discriminate the outcomes of patients with high- or low-α-fetoprotein level.Conclusion:The hepatic inflammation status which was evaluated by GGT/ALT, could predicate HCC prognosis perfectly.Additionally,GGT/ALT was also related to the tumor burden although it was an inflammatory indicator.Therefore,we suggested that the hepatic inflammatory status was correlated with HCC recurrence/metastasis, and GGT/ALT was a good parameter for us to select patients with high recurrence risks.Moreover,since GGT/ALT was conveniently available even for out-patients,it should be possible for us to judge the fluctuation of tumor burden and hepatic inflammation status,and further to modify the strategy of certain therapies.PartⅡ.The relevance of peritumoral inflammatory（related） cells to HCC prognosisPurpose:To invest the involvement of peritumoral inflammatory（related） cells in HCC recurrence/metastasis,and their relation to hepatic inflammation status.Methods & Results:Two independent and random cohorts of HCC patients underwent curative resection in Zhongshan Hospital,Fudan University,were enrolled as study populations:cohort A,from Fed 2002 to Nov 2005,n=130;cohort B,from Jan 2002 to Dec 2006,n=207.The messenger RNA（mRNA） levels of the functional genes in aHSC（ie,FAP,SPARC,and TNC）,quantitated by real-time quantitative polymerase chain reaction,and the density of peritumoral Foxp3+T-regulatory cells （Tregs） and CD68+ macrophages（Mφ）,assessed immunohistochemically in tissue microarray sections,were positively correlated with the density of peritumoral aHSC. The density（P=0.007 for recurrence-free survival[RFS]and P=0.021 for overall survival[OS]） and functional genes（FAP,P=0.001 for RFS;SPARC,P=0.007 for RFS and P=0.021 for OS） of peritumoral aHSC independently contributed to high recurrence or death rates,as did peritumoral Tregs or Mφ.In addition,both density and FAP expression of aHSC were positively related to peripheral GGT/ALT level（r =0.176 and 0.204,respectively;P=0.04 and 0.02,respectively）.Moreover. peritumoral aHSC were related to more early recurrences.It is important to note that the density of peritumoral aHSC,in combination with FAP and SPACR mRNA or density of Tregs and Mφ,might predict prognoses more effectively.In the context of cohort B,we found density of MC was positively associated with that of peritumoral Treg（r=0.353,P＜0.001） and peripheral GGT/ATL（r=0.146,P=0.036）.MC can only predicate recurrence independently;however,when combined with Treg,MC was independently associated with both OS and recurrence.Similar to aHSC.high MC related to more early intrahepatic recurrence rather than later ones.Elevated peritumoral MC density was also related to larger tumor size and poor encapsulations. Conlusion:（a） Peritumoral aHSC was associated with HCC prognosis both in cellular and in gene level,aHSC can aggravate hepatic inflammation via its functional genes including FAP and SPARC,and through its role accumulating Treg and Mφ. Therefore,aHSC should be a promising therapeutic target to modulate hepatic inflammation response and prevent recurrence/metastasis.（b） Peritumoral MC can facilitate HCC recurrence/metastasis on a similar fashion to aHSC.（c） The positive relationships between GGT/ALT and aHSC and MC suggested that both peritumoral pathological parameters and peripheral biochemistry parameters can reflect the hepatic inflammation properly.These two methods were related to each other.PartⅢ.The relevance of cyclooxygenase-2 to the invasion and metastasis of hepatoma cells and HCC prognosisPurpose:To estimate the involvement of COX-2 in HCC IFN-αtreatment,this study is carried out to evaluate the relationships between COX-2 and the invasion and prognosis of HCC.Method:Seven hepatoma cell lines with changed metastatic potency were used, in addition to 80 randomly selected HCC patients without later or extrahepatic metastasis.COX-2 expressions were assessed in cell lines with Western-blot,and in tumor samples with quantitive RT-PCR.Results:（a） Cellular COX-2 expressions were simultaneously decreased with the decrease in matastastic power;（b） The intratumoral COX-2 expression was independent of any clinicopathologic parameters;（c） The higher COX-2 expression was.the poorer prognosis would be.The hazard ratios were 1.916 for recurrence and 3.522 for death,respectively（P=0.025 and＜0.001,respectively）.Conclusion:High COX-2 expression may promote the metastasis of hepatoma cells and facilitate early intrahepatic recurrence.If intervention can inhibit COX-2 expression,its therapeutic effect should be improved.PartⅣ.The influence over the metastatic power of MHCC97-H from low-dose IFN-αalone and in combination with oxymatrine in vitroPurpose:To evaluate the therapeutic role of low-dose IFN-αalone and in combination with oxymatrine,and the involvement of COX-2 in IFN-αeffect.Methods:According to documents.1×10³IU/ml was defined as low-dose IFN-α, and MHCC97-H cell line was selected for in vitro study.There were four interventions:control group with no intervention,low-dose IFN-αonly group, oxymatrine only group and low-dose IFN-α+oxymatrine group.We compared the proliferation,apoptosis and metastasis power of MHCC97-H cells under different intervention.In addition,COX-2 expression was evaluated in each group.Results:（a） Low-dose 1FN-αcould not inhibit the proliferation and metastasis of MHCC97-H,even the intervention was prolonged.Also,it was invalid in inducing cell apoptosis.（b） Oxymatrine could inhibit cell proliferation to some extent,with an IC₅₀=44.5mg/ml,and 10mg/ml was chosen for further intervention in order to not mask the effect of combination.（c） In a time-dependent fashion,the combination could inhibit cell proliferation and metastasis power significantly,and also increase cell apoptosis significantly.（d） Low-dose IFN-αcould increase the COX-2 expression in a time-dependent fashion.However,when oxymatrine was added,the COX-2 expression shrunk significantly.Conclusion:The in vitro anti-tumor capability of low-dose IFN-αis demolished for a COX-2 inducing ability.Oxymatrine can retrieve and improve the anti-tumor power of low-dose IFN-αvia its role in reversing COX-2 expression.We proposed that low-dose IFN-αin combination with oxymatrine could result in synergetic anti-tumor effects.PartⅤ.In vivo impact of low-dose IFN-αalone and in combination with oxymatrine on the hepatic inflammation status and the growth and progression of orthotopic implantated MHCC97-H tumorsPurpose:To validate the anti-tumor effect and inflammation modulation power of low-dose lFN-αin combination with oxymatrine in vivo.Methods:IFN-αwas used at 1×10⁶IU/ml,0.1ml/d（s.c.） and oxymatrine at 60mg/kg·d（i.p.）.Male BALB/c nu/nu mice（20g） were divided into four subgroups: control group（Ⅰ,n=12）,low-dose IFN-αgroup（Ⅱ,n=8）,oxymatrine group（Ⅲ,n= 8） and combination group（Ⅳ,n=6）.All interventions were given for 62 days unless the mice were dead before being sacrificed.There were 3,2 and 2 mice dead before day 62 in control,low-dose IFN-αand oxymatrine group,respectively.Another four subgroups（n=12 in each subgroup） were simultaneously established to compare the life span.So.there were respectively 15,14,14 and 12 mice in control,low-dose IFN-α,oxymatrine and the combination group for life span comparison.Results:（a） In groupⅡ.the mice weights were lowest and the tumor volumes were largest;while the thorough opposite occurred in groupⅣ.（b） Among four groups,the lung metastasis incidence of groupⅣwas the lowest（P=0.011）. Low-dose IFN-αcan increase the number of lung metastasis significantly（P=0.018）. （c） Compared to groupⅠ,the life span of groupⅣwas significantly improved;the life span of groupsⅡandⅢwere similar to each other.（d） The GGT or GGT/ALT,which was positively related to the volume of tumors in each group,was significantly elevated in groupⅡ,while suppressed in groupⅣ.（e） Intratumoral and peritumoral aHSC density was increased in groupⅡ,and decreased in groupⅣ.（f） Intratumoral COX-2 expressions in groupsⅠandⅡwere much higher than those in groupⅢandⅣ.Conclusion:Low-dose IFN-αcould not inhibit the orthotopic growth, progression and lung metastasis of implanted tumors in vivo.In addition,the life span of this group was not prolonged significantly.Contrarily,groupⅣdisplayed profound orthotopic tumor inhibition,lung metastasis depression and life span expansion effects.We suggested the combination could modulate the hepatic inflammation and directly inhibit tumor cells effectively.The former was displayed by a marked decrease in aHSC density and GGT/ALT level which was positively associated with tumor volumes.The latter was displayed by a profound decrease in lung metastasis incidence and numbers.Importantly,we proposed that the suppression of COX-2 was the key for oxymatrine to incorporate with low-dose IFN-α.Conclusions1.The hepatic inflammation status is of prognostic significance for HCC.Evaluations of either peripheral GGT/ALT or peritumoral inflammatory（related） cells can reflect the hepatic inflammation status and predicate prognosis perfectly. The former is convenient to practice and can be used in therapeutic effect monitoring.The latter provides us with mechanical explanations and promising therapeutic targets.2.Low-dose IFN-αis insufficient in inhibiting tumor cells and modulating hepatitis.When combined with oxymatrine,the invasion and metastatsis of tumor cells can be well handled.COX-2 decrease is the key for the cooperation of low-dose IFN-αand oxymatrine.3.This combination can inhibit orthotopic implanted tumors progression via modulating the hepatic inflammation status.The novelty of this study1.For the first time,we demonstrated and reported that peripheral GGT/ALT was tightly associated with HCC prognosis.This ratio was a linkage between hepatic inflammation and tumor burden.More importantly,it was applicable in screening patients with high recurrence risks and monitoring the therapeutic effect.2.For the first time,we demonstrated and reported that peritumoral aHSC and MC were related to the HCC prognosis.These parameters provided us with promising therapeutic targets and were also helpful in screening patients with high recurrence risks.3.We found that low-dose lFN-αwas insufficient in inhibiting the invasion and metastasis of tumor cells,and in modulating hepatic inflammation.On the contrary,the combination of low-dose IFN-αand oxymatrine were effective in tumor cell inhibition and hepatic inflammation modulation.COX-2 decrease was the key for the synergetic effect of this combination.The potential application of this project1.Both peripheral GGT/ALT level and peritumoral aHSC or MC can well reflect the hepatic inflammation status.These parameters with prognostic power are of clinical utility in screening high recurrence risk patients and monitoring therapies.2.The combination of low-dose IFN-αand oxymatrine is a novel adjuvant therapy for HCC patients after resection.更多还原