【作者】 王磊；

【导师】 李延青；

【作者基本信息】 山东大学 ， 内科学， 2014， 博士

【摘要】 自2000年以来有关国家和地区根据医学的进步、新药的上市和参考全球最新循证医学的证据,分别制订了慢性乙型肝炎防治指南或共识,并不断更新。我国2010年更新的慢性乙型肝炎防治指南指出慢性乙型肝炎的治疗需要最大限度地长期抑制乙型肝炎病毒(HBV),其抗病毒治疗长期性是由HBV和疾病的特点所决定,患者随意停药可带来较为严重的临床后果。但是强调不能自行停药和长期治疗,并不意味着患者没有停药的机会,停药是患者的治疗需求。有调查显示：愿意接受长时间抗病毒治疗的患者比例低、长期治疗的经济压力大、依从性差。抗HBV治疗属病因治疗,应该存在有限疗程的特点和停药的可能性,故美国肝病学会(AASLD)、欧洲肝病学会(EASL)和亚太肝病学会(APASL)以及我国的肝病专家分别对核苷(酸)类似物治疗慢性乙型肝炎的疗程进行了规定。国内外慢性乙型肝炎防治指南虽然对核苷(酸)类似物治疗慢性乙型肝炎提出了治疗终点、规定了停药标准,如HBeAg阳性慢性乙型肝炎患者的基本停药原则是实现HBeAg血清转换且必须继续巩固治疗,HBeAg阴性慢性乙型肝炎患者的停药标准是实现HBsAg消失或达到一定的治疗时间,这为临床实践提供了参考；但针对国内外慢性乙型肝炎防治指南规定的停药标准,循证医学的证据尚不足。同时,我们在临床工作中也还发现部分接受核苷(酸)类似物治疗的HBeAg P日性慢性乙型肝炎患者,采用美国FDA批准的雅培(Abbott)公司生产HBV血清标志物(HBVM)定量检测方法,HBeAg已消失(甚至HBsAg消失),而抗-HBe迟迟不能出现。对这类患者的停药,国内外慢性乙型肝炎防治指南均未涉及,能否停药缺乏循证医学的资料,也有进行探索性研究的必要。为了提高核苷(酸)类似物抗病毒药物的治疗效果、减少停药后的病情复发、指导临床医生用药和为国内外慢性乙型肝炎防治指南的有关核苷(酸)类似物治疗疗程提供循证医学证据。自2001年以来,我们采用前瞻性研究对达到一定停药标准(参阅AASLD、EASL、APASL和我国慢性乙型肝炎防治指南)的HBeAg阳性和阴性慢性乙型肝炎患者停药后的疗效持久性进行了长达10年的随访研究并探讨了影响因素。研究方法一、研究对象本前瞻性研究队列共纳入279例应用核苷(酸)类似物停药的慢性乙型肝炎患者。①HBeAg转换组入组118例,男82例,女36例,中位年龄23(5-52)岁；其中,拉米夫定90例、阿德福韦酯13例、替比夫定13例、恩替卡韦2例。停药标准为总疗程≥18个月,HBV-DNA<103copies/ml、ALT复常和HBeAg血清转换≥6个月。②HBeAg消失组入组78例,男53例,女25例,年龄31(15～70)岁；其中,拉米夫定55例、阿德福韦酯14例、替比夫定6例、恩替卡韦3例。停药标准为总疗程≥18个月,HBV-DNA<103copies/ml、ALT复常和IHBeAg消失≥6个月。③HBeAg阴性组入组83例,男67例,女16例,年龄34(6-65)岁；其中,拉米夫定67例、阿德福韦酯14例、恩替卡韦2例。药标准为总疗程≥24月,HBV-DNA<103copies/ml、ALT复常达18个月以上。对拉米夫定停药后复发的其中24例慢性乙型肝炎患者,再次拉米夫定进行再治疗：并将既往研究队列匹配均衡的112例单一拉米夫定初治患者作为历史对照,进行M204I/V(YMDD)变异发生率的比较。二、检测指标及随访观察(一)HBV-DNA定量检测：采用核酸荧光定量聚合酶链反应(PCR)法,试剂由深圳匹基生物工程股份有限公司提供,检测下限为103拷贝/ml,仪器为美国应用生物系统公司产ABi7000实时PCR扩增仪。(二)HBV血清标志物(hepatitis B virus marker, HBVM)定量检测：采用化学发光微粒子免疫技术(chemiluminescent microparticle immunoassay,CMIA),应用美国雅培(Abbott)公司生产ARCHITECT i2000型免疫分析仪及原装试剂。HBeAg≤10S/CO为阴性,抗-HBe≤1.0S/CO为阳性。本研究对HBeAg阳性患者,将HBeAg阴性、抗-HBe<0.5S/CO定义为HBeAg血清转换,将HBeAg阴性、抗-HBe阴性或≥0.5S/CO定义为HBeAg消失。(三)HBV基因耐药M204I/V(YMDD)检测：采用巢式聚合酶链反应—限制性片段长度多态技术(nt-PCR-RFLP)。(四)核苷(酸)类似物抗病毒药物停药后的随访观察：于停药后1、2、3、4、6、9和12个月,此后每6个月分别进行肝功能、HBV-DNA定量检测；若复发则检测HBVM定量,未复发者则每12个月检测HBVM定量。在随访观察中,若HBV-DNA≥104copies/ml和／或ALT复升,2周内复查仍HBV-DNA≥104copies/ml则为复发,即研究终点；未复发者随访时间均≥12个月,最后一次随访作为研究终点。(五)拉米夫定停药复发后再治疗的患者,每3个月检测肝功指标、HBV-DNA定量,随访观察达2年。若HBV-DNA<103copies/ml后,再次HBV-DNA≥104copies/ml;或持续HBV-DNA≥T04copies/ml超过6个月则检测M204I/V(YMDD)变异。出现M204I/V(YMDD)变异即为研究终点。三、统计学处理应用SPSS for windows13.0统计软件整理、分析数据。计量资料服从正态分布的采用t检验或方差分析；不服从正态分布的采用采用秩和检验；计数资料采用x2检验或Fisher’确切概率法。累计复发率和M204I/V(YMDD)变异发生率的计算采用乘积限法(Kaplan-Meier法),累计复发率和累积变异率的比较采用时序检验(Log-rank test);影响停药后疗效持久性的相关多因素分析采用Cox比例风险模型(Proportional Hazard Model). P<0.05为差异有统计学意义。结果一、各组慢性乙型肝炎患者用药时间和停药后随访时间HBeAg转换组、HBeAg消失组和HBeAg阴性组的用药中位时间分别为28.5(18～99)、42(18～108)和30(24～87)个月,HBeAg消失组用药时间最长。停药后随访的中位时间(不包括复发者)在HBeAg转换组、HBeAg消失组和HBeAg阴性组分别为48(12-144)、48(12-129)和60(12-144)个月。二、各组慢性乙型肝炎患者停药后累计复发率和复发病例的复发时间停药后HBeAg转换组、HBeAg消失组和HBeAg阴性组累计复发率分别为1年(12个月)时16.9%、39.7%和47.0%,3年(36个月)时18.8%、43.0%和54.1%,5年(60个月)时24.1%、43.0%和61.1%,10年(120个月)为24.1%、43.0%和64.3%。HBeAg阴性组累计复发率最高,IHBeAg消失组次之,HBeAg转换组最低,时序检验(Log-rank test) x2=30.133,P<0.001。HBeAg转换组在停药后随访期间,共复发25例。从停药第2个月开始出现复发,最长的复发时间为60个月(5年)。在停药后3、6、12和24个月时,累计复发病例分别占总复发病例数的44.4%(11例)、60.0%(15例)、80.0%(20例)和88.0%(22例)。HBeAg消失组在停药后随访期间,共复发33例。从第1个月开始出现复发,最长的复发时间为36个月(3年)。1在停药后3、6、12和24个月时,累计复发病例分别占总复发病例数的69.7%(23例)、84.8%(28例)、93.9%(31例)和96.7%(32例)。HBeAg阴性组慢性乙型肝炎停药后随访期间共复发49例,在第1个月即有8例复发,最长的复发时间为96个月(8年)。在停药后3、6、12和24个月时,累计复发病例分别占总复发病例数的46.9%(23例)、53.1%(26例)、79.6%(39例)和87.6%(43例)。三、核苷(酸)类似物抗病毒药物停药后持久性的相关因素分析(一)停药后疗效持久性的相关单因素分析对HBeAg阳性组停药时的年龄等12个因素进行单因素分析,结果显示,HBeAg转换组停药时的年龄(t=4.25l,P<0.001)和用药时间(z=2.257,P=0.024)在复发组和持续有效组之间的差异具有统计学意义。HBeAg消失组仅停药时的年龄在复发组和持续有效组之间的差异具有统计学意义(t=2.208,P=0.030)：对HBeAg阴性组停药时的年龄等11个因素进行单因素分析,仅HBV-DNA消失时间在复发组和持续有效组之间的差异具有统计学意义(z=2.257,P=0.024)。停药时的年龄在持续有效组为30.4±13.4岁低于复发组的35.0±11.2岁；但差异无统计学意义(t=1.679,P=0.092)。(二)停药后疗效持久性的相关多因素分析将118HBeAg转换组慢性乙肝的停药时年龄、性别、用药时间,治疗基线时的ALT水平、AST水平、LgHBVDNA定量值、HBV-DNA消失时间、HBVDNA消失持续时间、HBeAg转换持续时间、既往抗病毒治疗史、抗病毒药物和乙肝家族史共12个因素纳入Cox-险比例回归模型进行多因素分析,得最优回归模型。结果表明,停药时的年龄与停药后复发有关,回归系数B为0.080,其95%可信区间为1.045-1.124；用药时间与停药后复发有关,回归系数B为-0.098,其95%可信区间为0.833-0.988。将78例HBeAg消失组慢性乙肝的停药时年龄等共12个因素纳入Cox风险比例回归模型进行多因素分析,得最优回归模型。结果表明,停药时的年龄与停药后复发有关,回归系数B为0.034,其95%可信区间为1.0055-1.066。将83HBeAgl阴性慢性乙肝的停药时年龄、性别、用药时间,治疗基线时的ALT水平、AST水平、LgHBVDNA定量值、HBV-DNA消失时间、HBVDNA消失持续时间、既往抗病毒治疗史、抗病毒药物和乙肝家族史共11个因素纳入Cox风险比例回归模型进行多因素分析,得最优回归模型。结果表明,停药时的年龄与停药后复发有关,回归系数B为0.029,其95%可信区间为1.003-1.056；HBVDNA阴转发生的时间与停药后复发有关,回归系数B为0.087,其95%可信区间为0.833-0.988。四、影响核苷(酸)类似物抗病毒药物停药后持久性因素的分层分析三组病例单因素和多因素分析均证明,停药时的年龄等因素是停药后复发的预测因素,对停药时年龄进行分层为≤20岁、21-30岁、21-30岁和>40岁4组。HBeAg转换组随停药时年龄的增大,停药后累计复发率升高,尤其30岁以上年龄组累计复发率升高明显,停药后5年(60个月)累计复发率高达50%以上,而30岁及以下年龄组累计复发率较低,停药后5年的累计复发率仅为10%左右。通过运用受试者工作特征曲线(ROC),计算ROC曲线下的面积(area under the ROC curve, AUC)来决定最佳年龄临界点。当年龄的临界点(cut-off值)为30.5岁时,ROC曲线下面积最大为0.77,得出30岁可作为最佳年龄分界线。将30岁作为年龄分界线分为2组计算停药后累计复发率,>30岁年龄组累计复发率高,停药后1、2-3、4和5年累计复发率分别约为35%、40%、450%0和55%；而<30则岁累计复发率约10%。HBeAg消失组同样随停药时年龄的增大,停药后累计复发率升高,尤其40岁以上年龄组累计复发率高,停药后6个月累计复发率高达50%以上。通过运用ROC,计算AUC。当年龄的临界点(cut-off值)为26.5岁时,ROC曲线下面积最大为0.62,得出26岁可作为最佳年龄分界线。将26岁作为年龄分界线分为2组,计算停药后累计复发率,结果显示,>26岁年龄组累计复发率高,停药后1年累计复发率接近50%,3年后未见复发。≤26岁年龄组累计复发率较低,停药后1年累计复发率约为25%,1年后未见复发。HBeAg阴性组结果显示,21岁以上年龄组累计复发率均高。当年龄的临界点(cut-off值)为20.5岁时,ROC曲线下面积最大为0.59,得出20岁可作为最佳年龄分界线。将20岁作为年龄分界线分为2组,计算停药后累计复发率,结果显示,虽HBeAg阴性组停药后累计复发率高；但≤20岁年龄组累计复发率较低,停药后1年累计复发率低于30%,1年后未见复发。而>20岁年龄组累计复发率高,停药后1年累计复发率51%,停药后2-3年接近60%,4年以上累计复发率约为70%,8年后未见复发。五、治疗和停药随访中HBsAg消失和HBsAg血清转换的情况停药时,HBsAg消失有9例,在HBeAg转换组、消失组和HBeAg阴性组分别有3例、4例和2例；其中HBsAg血清转换有3例,在HBeAg转换组、消失组和HBeAg阴性组分别有1例、1例和1例。在停药随访中,HBsAg消失累积至29例,在HBeAg转换组、消失组和HBeAg阴性组分别有9例、10例和10例,其中HBsAg血清转换已累积至19例,为HBeAg转换组7例、消失组6例、HBeAg阴性组6例。治疗和停药随访中出现HBsAg消失的患者未见停药复发现象。六、24例拉米夫定停药复发再治疗慢性乙型肝炎患者与作为历史对照的112例初治患者的M204I/V(YMDD)变异累计发生率,在用药后1年和2年分别为45.8%、17.9%和50.0%、32.6%。停药复发再治疗慢性乙型肝炎患者的M204I/V(YMDD)变异累计发生率高于初治患者(P=0.017)。结论及意义一、各组慢性乙型肝炎患者停药后累计复发率不同。HBeAg阴性组累计复发率最高,HBeAg消失组次之,HBeAg转换组最低,停药l0年的累计复发率分别为24.3%、43.0%和64.1%。二、各组慢性乙型肝炎患者复发病例的复发时间各有特点,应在停药后随访中加以注意。HBeAg转换组复发出现于停药后的2个月到60个月,5年后未见有复发；停药3-6个月时,复发病例约占到45%-60.0%,停药l-2年时,复发病例约占到80%-90.0%。HBeAg消失组复发出现于停药后的1个月到36个月,3年后未见有复发；复发病例主要集中在停药后1年内,停药3个月时,复发病例就占到接近70.0%,停药1年时,复发病例就占到接近94%。HBeAg阴性组停药后复发出现早,第1个月就复发8例,占复发总病例的16.3%；停药3-6个月时,复发病例占到50.0%左右；停药后8年内均有复发。三、核苷(酸)类似物抗病毒药物停药后持久性的相关因素分析。HBeAg阳性组单因素分析和多因素分析均表明停药时的年龄和用药时间与停药后复发有关,是停药后复发的独立预测因素。停药时的年龄偏大和用药时间短则复发率高,是停药后复发的高危因素。HBeAg消失组单因素分析和多因素分析均表明停药时的年龄与停药后复发有关,是停药后复发的独立预测因素。停药时的年龄偏大则复发率高,是停药后复发的高危因素。HBeAg阴性组多因素分析表明停药时的年龄和HBVDNA阴转发生的时间与停药后复发有关,是停药后复发的独立预测因素。停药时的年龄偏大和服用核苷(酸)类抗病毒药物后HV-DNA消失晚则复发率高,是停药后复发的高危因素。四、停药时年龄作为影响核苷(酸)类似物抗病毒药物停药后持久性的独立预测因素,通过运用ROC计算AUC得出的年龄临界点。HBeAg转换组将30岁作为最佳年龄年龄分界线,>30岁年龄组累计复发率高,停药后1、2-3、4和5年累计复发率分别约为35%、40%、45%和55%；而<30则岁累计复发率约10%。核苷(酸)类似物抗病毒药物停药时,一定要考虑到年龄因素,年龄大于30岁停药要慎重,应适当延长疗程。HBeAg消失组将26岁作为年龄分界线,>26岁年龄组累计复发率高,停药后1年累计复发率接近50%,3年后未见复发。≤26岁年龄组累计复发率较低,停药后1年累计复发率约为25%,1年后未见复发。年龄小于26岁,如患者有停药的意愿,可考虑停药并注意随访。HBeAg阴性组将20岁可作为年龄分界线,≤20岁年龄组累计复发率较低,停药后1年累计复发率低于30%,1年后未见复发。而>20岁年龄组累计复发率高,停药后1年累计复发率51%,停药后2～3年接近60%,4年以上累计复发率约为70%。年龄小于20岁,如患者有停药的意愿,可考虑停药并注意密切随访。五、治疗和停药随访中发生HBsAg消失的患者未见停药复发现象。对于高停药复发率的]HBeAg阴性和年龄偏大的HBeAg阳性慢性乙型肝炎患者,可考虑HBsAg肖失再停药。六、慢性乙型肝炎患者停药复发后再次拉米夫定单药治疗时仍有效；但耐药变异发生率明显高于初治患者。故对拉米夫定停药复发后再次治疗时,应选用耐药发生率低的核苷(酸)类似物或干扰素类抗病毒药物。更多还原

【Abstract】 Background and AimsSeveral practice guidelines on chronic hepatitis B (CHB) have been established by some national or regional associations based on the medical evidences and have been updated periodically since2000.As indicated in Chinese guideline (update2010), the overall objectives of CHB treatment are:To suppress hepatitis B virus replication to the most extent, reduce liver cell inflammation, necrosis and liver fibrosis, delay and reduce hepatic decompensation, liver cirrhosis, hepatocellular carcinoma and their complications, thereby improving the quality of life and prolong survival time. CHB require long-term antiviral therapy, which is determined by the characteristics of HBV and the disease, Arbitrary withdrawal of antivirals can bring more serious clinical consequences; however, the importance of long-term treatment and adverse sequences of arbitrary withdrawal do not necessarily mean that the antivirals cannot stop in any occasion.In fact, proper discontinuation of antiviral treatment is one of the patient’s demands. A survey shows that:fewer patients willing to accept a long-term antiviral therapy, long-term adherence to treatment is poor, longer treatment causes higher economic pressures. Antiviral therapy is an etiological treatment, and a finite treatment course or discontinuation of treatment might be possible. The antiviral treatment durations were defined in the guidelines from American Association for the Study of Liver Diseases (AASLD), the European Association for the Study of the Liver (EASL), the Asian Pacific Association for the Study of the Liver (APASL) and Chinese CHB guideline.Although antiviral cessation criteria were proposed by the guidelines, such as a certain consolidation period after HBeAg seroconversion in HBeAg-positive CHB patients or achievement of HBsAg loss in HBeAg-negative CHB patients, but evidence-based medicine data is still insufficient. Meanwhile, we also found a subgroup of HBeAg-positive CHB patients whose HBeAb remained negative despite of HBeAg loss or HBsAg loss (measured with Abbott reagents) in our clinical practice. The cessation criteria in this subgroup also warrants exploration.In order to improve the treatment efficacy of nucleos(t)ide analogues(NA), reduce relapse after discontinuation, provide a reliable reference for clinicians and EBM data for guidelines. A prospective follow-up study on the durability of efficacy in the HBeAg-positive and HBeAg-negative CHB who met certain cessation criteria (see AASLD, EASL, APASL and Chinese CHB guidelines) after NA withdrawal were conducted since2001to evaluate the off-treatment durability of NA treatment and explore the predicting factor for the relapse.Material and MethodsPatientsThis prospective cohort study included279CHB patients who discontinued NA.①118patients in the HBeAg seroconversion group,82males,36females, aged5to52years (median23); Among them,90were given with lamivudine,13with adefovir dipivoxil,13with telbivudine, and2with entecavir. Cessation criteria were as follows:the total withdrawal of treatment≥18months, HBV DNA<103copies/ml, ALT normalization and HBeAg seroconversion maintained≥6months.②HBeAg loss group enrolled78patients,53males and25females, aged15to70years (median31); Among them,55patients with lamivudine,14with adefovir dipivoxil,6patients with telbivudine,3patients with entecavir. Cessation criteria:the total withdrawal of treatment≥18months, HBV-DNA<103copies/ml, ALT normalization and HBeAg loss maintained for≥6months.③HBeAg-negative group enrolled83patients,67males and16females, aged6to65(median34) years of age;67patients with lamivudine,14patients with adefovir, two patients with entecavir. Cessation criteria:total treatment≥24months, HBV-DNA<103copies/ml, ALT normalization maintained≥18months. Among relapsers after LAM withdrawal,24patients were retreated with LAM (100mg/day); and another112patients with initial lamivudine treatment well matched with the relapsers were selected as historical control to compare the incidence of YMDD motif mutation (rtM204I/V).MethodsThe detecting parameters and follow-up(A) HBV DNA quantification was performed by real time PCR (reagents by PG biotech, Shenzhen, China; detection limit of1000copies/ml) using ABi7000real-time PCRthermocycler.(B) HBV serological markers (hepatitis B virus marker, HBVM)quantitative detection was performed with chemiluminescent microparticle immunoassay (CMIA) with Abbott ARCHITECT i2000immunoassay analyzer and the original reagents. HBeAg≤1.0S/CO was defined as negative, anti-HBe≤1.0S/CO as positive. In the present study, the negative HBeAg plus anti-HBe<0.5S/CO is defined as HBeAg seroconversion in HBeAg-positive patients; anti-HBe negative or≥0.5S/CO is defined as HBeAg loss.(C) HBV genetic resistance mutation rtM204I/V detection:restriction fragment length polymorphism technique after nested PCR (nt-PCR-RFLP) were used to detect the mutation.(D) Follow ups after discontinuation were performed at1,2,3,4,6,9and12months after treatment, and then every6months thereafter.Conventional physical examination, liver function and HBV DNA quantification were performed in every visit. HBVM quantification was performed immediately if relapse founded, or it was performed every12months. Relapse was defined as HBV-DNA≥104copies/ml (confirmed by another test2weeks apart) and/or ALT elevated. Relapse was the primary endpoint of this study. For non-relapsers (all were followed up for≥12months), last follow-up for the study endpoint.(E) For the relapsers retreated with lamivudine, follow-up visits were made every3months and liver function and HBV-DNA were tested, the total follow-ups duration were≥2years. The YMDD motif mutation (rtM204I/V)was tested if relapse was observed (definition see above) or HBV-DNA≥104copies/ml lasted more than6months. The appearance of YMDD motif mutation (rtM204I/V) was the primary endpoint in this section.Statistical analysisAll data were collected and input into SPSS for windows13.0statistical software to organize, analyze data. Measurement data were tested with t test or Wilcoxon test; count data using x2test or Fisher’s exact test. The cumulative relapse rate and M2041/V (YMDD) mutation rate were calculated and tested using Kaplan-Meier method and Log-rank test.Factors related with off-treatment efficacy were explored using Cox proportional hazards model (Proportional Hazard Model). P<0.05was considered statistically significant.ResultTreatment duration and follow-up timeThe treatment durations were18to99months (median28.5),18to108months (median42), and24to87months (median30) months in HBeAg seroconversion group, HBeAg loss and HBeAg-negative group respectively. The median follow-up time (excluding relapsers) in HBeAg seroconversion group, HBeAg loss group and HBeAg-negative group was48(12to144),48(12to129), and60(12to144) months, respectively.Cumulative relapse rate and relapse time after discontinuationThe cumulative relapse rates in HBeAg seroconversion group, HBeAg loss group and HBeAg-negative was16.9%,39.7%and47.0%within1year;18.8%,43.0%and54.1%within3years;24.1%,43.0%, and61.1%within5years;24.1%,43.0%and64.3%within10years. HBeAg-negative group had the highest cumulative relapse rate, followed by the HBeAg toss group and HBeAg seroconversion group (log-rank test (Log-rank test) x2=30.133, P<.001.25patients in HBeAg seroconversion group relapsed, the relapse time ranged from2months to5years, the cumulative relapse relapse were44.4%of total number of cases (11cases),60.0%(15cases),80.0%(20cases) and88.0%(22cases) at6,12and24months after treatment. 33patients in HBeAg loss group relapsed. the relapse time ranged from1months to36months.the cumulative relapse relapse69.7%respectively of the total number of cases (23cases),84.8%(28cases),93.9%(31cases) and96.7%(32cases) at6,12and24months after treatment.49patients in HBeAg loss group relapsed, the relapse time ranged from1months to8years.The cumulative relapse relapse46.9%respectively of the total number of cases (23cases),53.1%(26cases),79.6%(39cases) and87.6%(43cases) at6,12and24months after treatment.Factors related with off-treatment efficacy durability of NA treatment(A) Univariate analysisIn HBeAg seroconversion group,12variables were analyzed and the results showed that age at withdrawal (t=4.251, P<0.001) and the treatment time (z=2.257, P=0.024) differedbetween relapsers and non-relapsers. In HBeAg loss group, only age at withdrawal diffed between relapsers and non-relapsers (t=2.208, P=0.030); In HBeAg-negative group,11variables were analyzed and only HBVDNA loss time between relapsers and non-relapsers was statistically different (z=2.257, P=0.024). Age at withdrawal of non-relapsers was30.4±13.4years and that of relapsers was35.0±11.2years; but the difference was not statistically significant (t=1.679,P=0.092).(B) Multivariate analysisThe HBeAg seroconversion group when118cases of chronic hepatitis B HBeAg withdrawal age, sex, time of administration, ALT levels at baseline treatment, AST levels, LgHBVDNA quantitative values, HBV-DNA disappearance time, HBVDNA disappear duration, HBeAg conversion duration, previous antiviral treatment history, family history of hepatitis B antivirals and12factors into the Cox proportional hazards regression model for multivariate analysis, get the best regression model. The results showed that when age and withdrawal relapse after withdrawal regression coefficient B is0.080,95%CI1.045-1.124; administration time and relapse after treatment, the regression coefficient B is-0.098,95%CI0.833-0.988.When the group of78cases of chronic hepatitis B HBeAg loss withdrawal age of 12factors into the Cox proportional hazards regression model for multivariate analysis, was the best regression model. The results showed that when age and stopping relapse of withdrawal, regression coefficient B is0.034,95%confidence interval,1.0055-1.066.The83cases when HBeAg negative chronic hepatitis B withdrawal age, sex, time of administration, ALT treatment baseline levels, AST levels, LgHBVDNA quantitative values, HBV-DNA disappearance time, HBVDNA disappear duration, previous treatment history of anti-virus, anti-hepatitis B viral drugs and family history of11factors into the Cox proportional hazards regression model for multivariate analysis, get the best regression model. The results showed that when age and withdrawal relapse after withdrawal, regression coefficient B of0.029,95%CI1.003-1.056; time HBVDNA seroconversion occurred with relapse after treatment, the regression coefficient is0.087B,95%confidence interval0.833-0.988.Stratified analysis on the impact factors of off-treatment durabilityThree patients were univariate and multivariate analyzes were performed to prove, age and other factors withdrawal when the predictors of relapse after withdrawal, the withdrawal of the age when stratified as≤20years old,21to30years old,21to30years old and>40years4groups.When HBeAg conversion with withdrawal age group increases, the cumulative relapse rate increased after treatment, especially in the age group over30years of cumulative relapse rate increased significantly, five years after discontinuation (60months) cumulative relapse rate of50%or more, and30-year-old age group and the cumulative relapse rate is low, the cumulative5-year relapse rate after treatment is about10%. By using receiver operating characteristic curve (ROC), calculate the area under the ROC curve (area under the ROC curve, AUC) to determine the optimal age of the critical point. When the critical point of age (cut-off value) was30.5years old, the largest area under the ROC curve of0.77, can be obtained as the best30years of age boundaries. The30-year-old as the age of the line is divided into two groups to calculate the cumulative relapse rate after treatment,>30age group the cumulative relapse rate after discontinuation of1,2-3,4, and5-year cumulative relapse rate was approximately35%,40%,45%and55%;<30years of age and the cumulative relapse rate of about10%.HBeAg disappears when the same group with the withdrawal age increases, the cumulative relapse rate increased after treatment, especially in the age group over40years of cumulative relapse rate, six months after stopping the cumulative relapse rate of over50%. Through the use of ROC, calculate AUC. When the critical point of age (cut-off value) was26.5years old, the largest area under the ROC curve was0.62, obtained26years as the best age boundaries. The26-year-old as the age of the line is divided into two groups, calculated the cumulative relapse rate after treatment, the results showed that>26age group the cumulative relapse rate after stopping a cumulative relapse rate of nearly50%after3years and no relapse.<26years age group the cumulative relapse rate is low, after stopping a cumulative relapse rate of about25%, and no relapse after1year.HBeAg-negative group showed that the age group over21years cumulative relapse rate is high. When the critical point of age (cut-off value) was20.5years old, the largest area under the ROC curve was0.59, obtained20-year-old as the best age boundaries. The20-year-old as the age of the line is divided into two groups, calculated the cumulative relapse rate after treatment, the results showed that the cumulative relapse rate after discontinuation, although HBeAg-negative group;<20years age group, but lower cumulative relapse rate after stopping1-year cumulative relapse rate of less than30%, no relapse after1year. And>20age group the cumulative relapse rate after stopping a cumulative relapse rate of51%,2to3years after stopping nearly60%, more than four-year cumulative relapse rate of about70%, and no relapse after8years.HBsAg loss and HBsAg seroconversion during treatment and follow upWhen stopping, HBsAg disappears in9cases, in HBeAg seroconversion group, disappeared group and HBeAg-negative group,3cases,4cases and2cases, respectively; among HBsAg seroconversion have3cases, in HBeAg seroconversion group, disappeared group and HBeAg-negative group were1case,1 case and1cases.In follow-up withdrawal, HBsAg disappears accumulated to29cases in HBeAg seroconversion group, disappeared group and HBeAg-negative group were9cases,10cases and10cases; in which HBsAg seroconversion has accumulated to19cases for HBeAg seroconversion group7cases,6cases of disappearance group, HBeAg-negative group6cases. Discontinuation of treatment and follow-up of patients with HBsAg loss occurred in the no stopping relapse.Resistance mutation in lamivudine retreated patients and naive patients24elapse after discontinuation of lamivudine treatment of chronic hepatitis B patients with a history of112patients with previously untreated control patients with M204I/V(YMDD) mutation cumulative incidence after treatment1and2years, respectively45.8%,17.9%and50.0%,32.6%. Discontinuation of treatment of chronic hepatitis B relapse in patients with M204I/V(YMDD) mutation cumulative incidence in previously untreated patients (P=0.017).Conclusion and SignificanceFirst, The cumulative relapse rates after discontinuation in various groups are different. HBeAg-negative group has the highest cumulative relapse rate, followed by HBeAg loss group and HBeAg seroconversion group. The cumulative relapse rates within10years were64.3%,43.0%and24.1%, respectively.Second, The relapse time in different groups has their own characteristics, which should be noted in the follow-up after treatment. HBeAg conversion group relapsed in two months after stopping to60months, no relapse after5years; withdrawal from3to6months, relapse accounted for about45%to60.0%, withdrawal1-2years, relapse accounted to80%to90.0%. HBeAg loss group relapsed after discontinuation of one month to36months, no relapse after3years; relapse mainly within one year after the withdrawal, while stopping three months, relapse accounted for nearly70.0%, withdrawal1year, relapse accounted for nearly94%. HBeAg-negative group relapsed after stopping early, the first months of relapse in8cases, accounting for16.3%of total recurrent cases; withdrawal from3to6months, relapse accounted for about50.0%; relapse can occur within as long as eight years.Third, Related factors with off-treatment durability of antiviral drug.For HBeAg seroconversion group, univariate analysis and multivariate analysis showed that when age and total treatment duration were independent predictors of relapse after treatment. Patients with older age and short treatment duration are more likely to relapse.For HBeAg loss group, univariate analysis and multivariate analysis showed that the age at withdrawal was an independent predictor of relapse after treatment. When the older withdrawal relapse rate, is a risk lactor of relapse after treatment.For HBeAg-negative group, multivariate analysis showed that the age and HBVDNA loss time were independent predictors of relapse after treatment. Older age and late HBV-DNA loss are high risk factors of relapse after treatment.Fourth, Age as independent predictors of off-treatment efficacy, and the cutoff value o fage using ROC analysis.The cutoff value of age in HBeAg seroconversion group is30years, the cumulative relapse rate in patients>30years after discontinuation was approximately35%,40%,45%and55%at of1,2-3,4and5years; in patients≤30years old the cumulative relapse rate of about10%. When considering NA withdrawal, age at withdrawal must be taken into account. Patients with older than30years should prefer extended treatment rather than N A withdrawal.The cutoff value of age in HBeAg loss group is26years, the cumulative relapse rate in patients>26years after discontinuation was approximately50%within1year; in patients≤26years old the cumulative relapse rate of about25%. When considering NA withdrawal, age at withdrawal must be taken into account. NA withdrawal could be considered in patients with≤26years, and should be carefully followed up.The cutoff value of age in HBeAg-negative group is20years, the cumulative relapse rate in patients≤20years after discontinuation was less than30%within1year; in patients>20years old the cumulative relapse rate of about51%within1year, and up to60%within2-3years. NA withdrawal could be considered in patients with≤20years, and should be carefully followed up.Fifth, No relapses were observed in all patients who experienced HBsAg loss in follow up. For patients with high risk of relapse in HBeAg-negative group and HBeAg-positive group, NA can be stopped when HBsAg disappear.Sixth, Lamivudine re-treatment is still effective for relapsers after lamivudine withdrawal; but resistance mutations are significantly more frequent than untreated patients. Therefore, NA with low resistance rates or interferon should be preferred when considering re-treatment in spite of lamivudine.更多还原