作者：唐立成
美工：何国红 罗真真
排版：马超
2025年5月的一个清晨，基因泰克的传奇人物Paul Carter博士坐在办公室里，指尖划过Antibody Society最新更新的表格-全球获批的抗体治疗药物已达212个。他想起半个世纪前，自己还是个研究生时，导师Greg Winter递来的那篇《自然》论文：1975年，Georges Köhler和César Milstein用小鼠B细胞和骨髓瘤细胞融合，造出了能精准结合特定抗原的“单克隆抗体”。而这项基于杂交瘤的技术发明，也在9年后的1984年，帮助Georges Köhler和César Milstein与Niels Jerne共享诺贝尔奖。
那时没人能想到，这个在实验室里诞生的“小工具”，会在五十年后长成参天大树-年销售额超2000亿美元，拯救了数百万癌症、自身免疫病患者的生命。这五十年里，有实验室里的彻夜不眠，有临床试验的跌宕起伏，有企业间的技术博弈，更有患者重获新生的热泪盈眶。
这不是一串冰冷的技术名词，而是无数医药人用青春和坚守写就的故事。
▶ 楔子：1975年的“意外之喜”
1974年冬天，英国剑桥MRC实验室里，Köhler和Milstein对着一堆失败的细胞培养皿发愁。他们想解决一个困扰免疫学界多年的难题：如何得到“纯”的抗体-不是血清里混杂的多种抗体，而是能精准瞄准一个抗原的“独苗”。
▲ 生物化学家César Milstein（左）和Georges Köhler不仅是实验室的战友也是生活的密友
此前，科学家们尝试过各种方法，都以失败告终。直到Milstein突然想到：骨髓瘤细胞能无限分裂，而B细胞能产生特异性抗体，要是能把两者“拼”在一起呢？
他们的操作听起来简单，却藏着无数细节：先把绵羊红细胞（一种简单的抗原）乳化后注射进小鼠体内，等小鼠脾脏里的B细胞被激活，再取出这些能分泌抗体的B细胞，和不能产生抗体的骨髓瘤细胞混合，加入聚乙二醇（PEG）-这种像糖浆一样的化学物质，能让细胞膜温柔地融合。然后，他们把混合细胞放进加了选择培养基的培养皿里，一天天观察哪些细胞能存活、能分泌抗体。
▲ 杂交瘤技术原理
几周后的一个早晨，Köhler发现某个培养孔里的细胞不但活得好好的，还分泌出了只结合绵羊红细胞的抗体。“我们成功了！”他冲进Milstein的办公室，手里攥着检测报告，声音都在发抖。
1975年8月，《自然》发表了这篇仅两页的论文。当时没人意识到，这篇论文会开启一个全新的治疗时代-9年后的1984年，Köhler和Milstein与提出“免疫系统特异性理论”的Niels Jerne共享诺贝尔生理或医学奖。领奖台上，Milstein笑着说：“我们只是想做个研究工具，没想到它会变成拯救生命的药物。”
▲ 1975年在自然杂志发表的仅2页的“杂交瘤技术”问世论文
更珍贵的是，这个“工具”没有被锁在实验室里。论文发表后，Milstein实验室的电话快被打爆了-全球的研究者都来索要杂交瘤细胞系。他们没要一分钱，把这些“科学种子”装进装满干冰的盒子，寄往美国、日本、德国的实验室。有学者后来回忆：“我收到的细胞系里，还夹着一张手写的纸条，上面写着‘注意培养温度，有问题随时来信’。”
当时有立法者批评MRC“不懂知识产权保护”，但如今回头看，这种开放恰是抗体产业的起点。要知道，现代生物技术产业那时还在萌芽，英美高校直到多年后才被允许为政府资助的科研成果申请专利。这些科学家的初衷，只是想解答“抗体如何产生”的基础问题，却在无意间为千万患者埋下了希望的种子。
半个世纪里，抗体领域的光芒从未熄灭：至少有六位科学家因抗体药物开发的突出贡献，先后获得诺贝尔奖或拉斯克奖-从发现抗体多样性遗传基础的Susumu Tonegawa，到开创免疫检查点疗法的James Allison，每一个奖项背后，都是对Köhler和Milstein最初探索的传承。
▶ 第一章：鼠源抗体的“困境与突围”
杂交瘤技术诞生后，科学家们像发现了新大陆。很快，小鼠来源的单克隆抗体成了实验室里的“明星工具”-能精准识别细胞表面的蛋白，帮科研人员找到癌症、感染的靶点。但当大家想把它变成治疗药物时，却撞上了“南墙”。
1986年，全球首个抗体药muromonab-CD3获批，用于预防手术后的排斥反应。这是个鼠源抗体，看起来是个好开始，可临床应用却给了所有人一记耳光：患者注射后，身体把它当成“外来入侵者”，产生抗药抗体（ADA），短短几天就把药物清除掉，疗效越来越差；有的患者出现严重过敏，甚至休克；更麻烦的是，鼠源抗体在人体内的半衰期只有1-2天，患者得频繁输液，痛苦不堪。当时负责muromonab-CD3临床的医生回忆：“有个手术患者，用了两周后，抗体完全失效，排斥反应差点夺走他的生命。我们看着明明有效的药物，却没办法让它在体内多待一会儿。
”这就是鼠源抗体的“三宗罪”：免疫原性强、半衰期短、效应功能弱。要让抗体药真正走进临床，必须“改头换面”-把小鼠的抗体，变成更像人类自身的抗体。
第一个破局的是“嵌合抗体”。1984年，两个团队同时想到：把小鼠抗体的“识别部分”（可变区），嫁接到人类抗体的“骨架部分”（恒定区）。这样一来，抗体里人源序列占67%，鼠源序列只剩33%，免疫原性大大降低。
1994年，首个嵌合抗体阿昔单抗（abciximab）获批，用于预防心脏支架术后的血栓。它能精准结合血小板表面的GPIIb/IIIa受体，阻止血小板聚集。临床数据显示，它能把血栓风险降低30%，而且过敏反应比muromonab-CD3少了一半。
但科学家们不满足。1986年，英国剑桥的Winter博士团队又往前迈了一步-“人源化抗体”：只把小鼠抗体里负责识别抗原的6个小loops（CDR区）移植到人源抗体上，人源序列占比高达90%。就像给小鼠抗体换了件“全人类的衣服”，只留下识别靶点的“眼睛”。
1997年，首个人源化抗体达利珠单抗（daclizumab）获批，用于手术抗排斥。这次，患者的抗药抗体发生率降到了10%以下，半衰期延长到了14天。医生们终于不用再频繁调整剂量，患者也不用再承受过敏的痛苦。
而真正的“里程碑”，出现在2002年。首个全人源抗体阿达木单抗（adalimumab）获批治疗类风湿关节炎-它不是改造鼠源抗体，而是用噬菌体展示技术，直接从人类抗体库中筛选出来的，100%人源序列。
▲ 初代阿达木单抗“修美乐”
阿达木单抗的研发负责人回忆：“我们筛选了上百万个噬菌体，才找到这个能精准结合TNF（肿瘤坏死因子）的抗体。临床实验里，有个患者关节肿痛得没法走路，用了三个月后，居然能正常爬山了。”后来，阿达木单抗的适应症扩展到9种自身免疫病，巅峰年销售额超200亿美元，成了首个“年销20亿俱乐部”的抗体药。
从鼠源到嵌合，再到人源化、全人源，这一步走了27年。背后是无数次的基因拼接、蛋白表达、动物实验-就像在黑暗里摸路，每一步都充满未知，却又朝着“让药物更安全、更有效”的方向坚定前行。
▶ 第二章：技术分流：ADC与双抗的“江湖纷争”
当全人源抗体在自身免疫病领域“大杀四方”时，癌症治疗领域的科学家们却在思考：抗体能不能更“狠”一点？只结合靶点还不够，能不能带个“毒药”，精准杀死癌细胞？
这就是抗体药物偶联物（ADC）的想法-把抗体当成“导弹”，把细胞毒性药物当成“弹头”，linker（连接子）就是“引线”，让导弹只在肿瘤细胞里引爆弹头。
这个想法早在上世纪70年代就有了，但直到2000年，首个ADC药物吉妥珠单抗ozogamicin（Mylotarg）才获批，用于治疗急性髓系白血病（AML）。它用抗CD33抗体做导弹，把能断裂DNA的卡奇霉素当弹头。
可命运给了它一记重击。2010年，FDA要求它撤市-因为后续临床试验发现，它会导致严重的肝损伤，甚至引发致命的肝静脉闭塞病。“当时整个ADC领域都凉了，大家都说这技术行不通。”参与研发的科学家后来回忆。
但总有不服输的人。基因泰克的团队没有放弃，他们重新设计了linker：以前的linker在血液里就容易断裂，导致弹头提前释放；新的linker像“密码锁”，只有进入肿瘤细胞的溶酶体，被特定酶切开才会释放弹头。
2013年，曲妥珠单抗emtansine（T-DM1）获批，用于HER2阳性乳腺癌。它用曲妥珠单抗（赫赛汀）做导弹，把抑制微管的美登素当弹头，DAR值（药物抗体比）控制在3.5-这个数值经过无数次实验验证，既能保证疗效，又不会有太多毒性。
临床数据让所有人振奋：以前用过赫赛汀和化疗的患者，用T-DM1后，肿瘤进展时间从4.6个月延长到9.6个月，而且副作用比化疗小太多。有个患者回忆：“以前化疗掉光了头发，吐得没法吃饭，用了T-DM1后，除了有点疲劳，基本能正常生活。”
更颠覆的是2019年获批的德曲妥珠单抗deruxtecan（T-DXd）。它的弹头是拓扑异构酶I抑制剂DXd，linker能精准在肿瘤里断裂，而且DXd能穿透细胞膜-就算旁边的癌细胞没表达HER2，也会被“误伤”，这就是“旁观者效应”。
2022年，T-DXd甚至获批用于“HER2低表达”乳腺癌-以前这类患者没法用HER2靶向药，现在T-DXd让他们有了新希望。研发团队的负责人说：“我们花了10年优化linker和弹头，光DAR值就测试了从2到8的各种组合，终于找到了最优解。”
▲ 抗HER2抗体治疗药物在肿瘤学中的应用
就在ADC药物“浴火重生”时，另一类抗体技术也在悄然崛起-双特异性抗体（双抗）。它能同时结合两个靶点，就像“双手”，一边抓着癌细胞，一边抓着免疫细胞，让免疫细胞“手撕”癌细胞。
2009年，首个双抗catumaxomab获批，用于治疗恶性腹水。它一边结合癌细胞表面的EpCAM，一边结合T细胞表面的CD3，把T细胞“拉”到癌细胞身边。但它是鼠源双抗，免疫原性强，后来因为商业原因撤市了。
真正让双抗“火起来”的是blinatumomab（贝林妥欧单抗）。2014年，它获批用于急性淋巴细胞白血病（ALL），是个单链双抗（BiTE格式），一边抓CD19（B细胞标志物），一边抓CD3。
它的半衰期只有2小时，患者需要连续静脉输液4周。但疗效惊人：以前复发的ALL患者，用博纳吐单抗后，完全缓解率从25%升到44%。有个16岁的患者，化疗失败后用了博纳吐单抗，现在已经无癌生存5年了。
后来，双抗技术越来越成熟：2022年获批的glofitamab用了“2+1”格式，两个臂抓CD20，一个臂抓CD3，亲和力更强；2023年获批的zenocutuzumab能同时抓HER2和HER3，解决了NRG1融合癌症的治疗难题。
▲ 临床试验中的常见双抗作用模式及抗原靶点
ADC和双抗，就像抗体治疗江湖里的“两大流派”-一个靠“精准导弹”，一个靠“借力打力”，却都朝着同一个目标：让癌症治疗更精准、更安全。
▶ 第三章：临床回响: 从“无药可医”到“带癌生存”
2004年，28岁的Lisa被诊断为HER2阳性转移性乳腺癌。当时医生告诉她：“化疗能延长一点时间，但大概率活不过3年。”
她参加了曲妥珠单抗（赫赛汀）的临床试验。第一次输液后，她紧张得整夜没睡，生怕有副作用。没想到，除了有点发热，什么不适都没有。3个月后复查，CT显示她肺部的转移灶缩小了30%；1年后，转移灶几乎消失。现在，Lisa已经52岁，是两个孩子的母亲，每年定期复查，癌症再也没复发过。
▲ 基因泰克抗体疗法患者Lisa的故事
曲妥珠单抗的故事，是抗体治疗改变癌症患者命运的缩影。1998年，它获批用于HER2阳性乳腺癌时，HER2阳性患者的5年生存率只有40%；现在，随着曲妥珠单抗、帕妥珠单抗、T-DM1、T-DXd的陆续获批，这个数字已经升到85%。
还有淋巴瘤患者Mike。2018年，他被诊断为弥漫大B细胞淋巴瘤，化疗后复发，医生说“没什么好办法了”。后来他参加了polatuzumab vedotin（CD79b ADC）的临床试验。这个药把抗CD79b抗体和微管抑制剂偶联，能精准杀死B细胞淋巴瘤细胞。
治疗6个月后，Mike的淋巴瘤完全缓解。他说：“以前化疗的时候，我连端杯子的力气都没有，现在能每周打两次篮球。我最感谢的，是那些在实验室里反复测试药物的科学家-是他们给了我第二次生命。”
不仅是癌症，抗体治疗也改变了自身免疫病患者的生活。45岁的Sarah患类风湿关节炎10年，关节畸形得没法系鞋带，吃了很多药都没用。2019年，她开始用阿达木单抗，每周皮下注射一次。现在，她不仅能正常做家务，还能陪女儿去跳舞。“以前我觉得自己是家里的累赘，现在我能照顾家人了。”
这些患者的故事，不是偶然。五十年里，每一个抗体药的获批，背后都是上万例临床试验数据的支撑；每一次疗效的提升，都是科学家们对“如何让药物更精准”的不断探索-比如优化抗体的Fc区，让它更能激活免疫细胞（ADCC效应）；比如设计“回收抗体”，让药物在体内停留更久（利用FcRn受体）；比如开发皮下注射剂型，让患者不用再频繁去医院输液。
就像Paul Carter博士说的：“我们做抗体治疗，不是为了发表论文，也不是为了拿奖项-我们的目标很简单，就是让患者能像正常人一样生活。”
▶ 第四章：幕后英雄：生物制造的“革命”
很少有人知道，一个抗体药从实验室走到病床，除了研发，还有个“卡脖子”的环节-生物制造。
1986年，muromonab-CD3获批时，生产1克抗体需要几十升细胞培养液，而且纯度很难达标。当时基因泰克的工程师回忆：“我们用不锈钢生物反应器，每次培养都要消毒好几天，产量却只有1-2g/L。有时候一批药因为纯度不够，只能全部倒掉，大家都心疼得掉眼泪。”
改变始于CHO细胞（中国仓鼠卵巢细胞）。这种细胞能高效表达人类蛋白，而且容易培养。1990年代，科学家们开始用基因工程改造CHO细胞，让它能更好地表达抗体。到2000年，CHO细胞的抗体产量达到5-8g/L；现在，经过优化的CHO细胞，产量能超过10g/L-相当于以前10升培养液的产量，现在1升就能完成。
▲ 中国仓鼠卵巢细胞（目前最广泛用于抗体生产的CHO细胞系）
还有“一次性生物反应器”的出现。以前用不锈钢反应器，每次使用后都要彻底清洗、消毒，既费时间又容易污染；一次性反应器即用即扔，不仅降低了污染风险，还能灵活调整产量。现在，全球70%以上的抗体药都是用一次性反应器生产的。
▲ 一次性生物反应器
生物制造的进步，不仅提高了产量，还降低了成本。1990年代，1克抗体药的生产成本要上万美元；现在，成本已经降到了几百美元。这也是为什么越来越多的抗体药能进入医保，让更多患者用得起。
“以前我们觉得，生物制造只是‘生产环节’，不重要。后来才发现，没有制造的进步，再好的研发成果也没法惠及患者。”一位生物制造工程师说。他曾为了优化培养基，连续一个月住在实验室，每天测试不同的营养成分比例，最后终于让CHO细胞的产量提高了20%。
▶ 第五章：未来：AI与“下一个五十年”
2024年，诺贝尔化学奖颁给了AlphaFold的研发团队-这个AI工具能根据氨基酸序列预测蛋白质结构，精度远超以前的方法。消息传来，抗体领域的科学家们都很兴奋：“以前我们设计抗体，要靠试错，现在有了AI，能直接在电脑上模拟抗体和靶点的结合，效率提高了10倍。”
比如研发“口服抗体”-这是科学家们的“终极梦想”。以前抗体只能注射，因为口服后会被胃酸和蛋白酶分解。现在，用AI设计的单域抗体（VHH），能抵抗蛋白酶的降解，而且容易穿透肠道黏膜。2025年，首个口服抗IL-23受体VHH抗体在猴子身上实验成功：一次口服后，能持续抑制IL-23信号24小时，和注射效果一样。
还有“穿越血脑屏障”的抗体。以前，抗体很难进入大脑，没法治疗阿尔茨海默病、脑肿瘤等疾病。现在，科学家们设计了“双抗”-一边结合靶点（比如淀粉样蛋白），一边结合转铁蛋白受体（TfR），能通过TfR介导的转运进入大脑。2024年，这类双抗在阿尔茨海默病的临床试验中，能显著减少患者大脑里的淀粉样斑块。
更让人期待的是“多特异性抗体”-能同时结合3个甚至更多靶点。比如同时结合PD-1、CTLA-4和LAG-3（三个免疫检查点），能更有效地激活免疫细胞，治疗难治性癌症。现在，已有多个三特异性抗体进入临床实验，初步数据显示疗效比双抗更好。
▲ 抗体药物发展50年风雨江湖
Paul Carter博士说：“五十年前，我们连‘人源抗体’都不敢想；现在，我们在探索‘口服抗体脑穿透抗体’。下一个五十年，抗体治疗可能会更神奇-比如能根据患者的基因定制抗体，或者能自动识别肿瘤的‘智能抗体’。”
▶ 尾声：五十年的传承，一场未完的“长征”
2025年6月，Köhler和Milstein的学生们在剑桥举办了一场纪念会。会上，有人展示了1975年那批杂交瘤细胞的冻存管-上面还贴着手写的标签：“绵羊红细胞抗体，1975.8.7”。
五十年里，抗体治疗从一个实验室里的“小工具”，变成了拯救千万人的“生物导弹”；从鼠源抗体的困境，到ADC、双抗的百花齐放；从不锈钢反应器的低产量，到一次性反应器的高效生产……每一步都充满挑战，每一步都离不开无数医药人的坚守。
有在实验室里熬夜做Western blot的研究生，有在病床前记录数据的医生，有在生产车间优化工艺的工程师，还有勇敢参加临床试验的患者-他们都是这场“长征”的参与者。
现在，年轻的科学家们还在继续前行：在显微镜前观察抗体和细胞的结合，在电脑上用AI设计新的抗体结构，在车间里测试新的制造工艺……他们知道，抗体治疗的故事还没结束，下一个五十年，还有更多的“不可能”等着被打破。
就像那句说的：“医学的进步，不是靠一个人的灵光一现，而是靠一代又一代人的接力。”抗体治疗的五十年，是这样；未来的五十年，也会是这样。
这是他们的故事，也是我们的故事-因为每一个技术突破的背后，都是“让患者更健康”的初心。
▶ 致谢：
在撰写这篇关于抗体治疗五十年历程的推文时，我始终怀着一份对“源头”的敬畏-这份创作的初心，完全源于一篇照亮思路的重磅综述：今年8月发表于《Nature Reviews Immunology》、由基因泰克三位资深科学家 Andrew C. Chan、Greg D. Martyn 与 Paul J. Carter 共同执笔的《Fifty years of monoclonals: the past, present and future of antibody therapeutics》。
我希望将复杂的技术迭代转化为普通人能理解的故事：以时间为轴，将单克隆抗体技术五十年的演进讲成一段有温度的“科学史诗”：从1975年杂交瘤技术在实验室的萌芽，到鼠源抗体的困境、嵌合/人源化/全人源抗体的突围，再到ADC、双抗、CAR-T等技术如何一步步打破治疗边界，最终开启现代抗体疗法的黄金时代。
Sanyou 10th Anniversary: From Dismissed to Lifesaving: 50 Years of Trials and Triumphs in Antibody Therapeutics
On a May morning in 2025, Genentech legend Dr. Paul Carter sat in his office, scrolling through the latest update from the Antibody Society: 212 antibody therapeutics approved worldwide. His mind drifted back half a century, to his days as a graduate student, when his mentor, Greg Winter, handed him a Nature paper. Published in 1975, it described how Georges Köhler and César Milstein fused mouse B cells with myeloma cells to create “monoclonal antibodies” that could bind precisely to a specific antigen. This hybridoma breakthrough earned Köhler and Milstein, together with Niels Jerne, the Nobel Prize in 1984.
At the time, no one could have imagined that this “lab tool” would grow into a towering tree-an industry worth more than $200 billion annually, bringing new treatment possibilities to patients with cancer and autoimmune diseases. Over these fifty years, there have been sleepless nights in the lab, the ups and downs of clinical trials, technological battles between companies, and, most moving of all, tears of joy from patients given a second chance at life.
This is not just a string of technical terms, but a story written with the youth, perseverance, and devotion of generations of biomedical pioneers.
▶ Prologue: The “Accidental Breakthrough” of 1975
In the winter of 1974, at the MRC Laboratory in Cambridge, Köhler and Milstein stared at rows of failed culture plates. They sought to solve one of immunology’s biggest challenges: how to obtain pure antibodies-not the polyclonal mix in serum, but a single antibody that could target one antigen with precision.
▲ César Milstein (left) and Georges Köhler were not only comrades in the lab but also close friends in life
Scientists had tried many approaches before, all without success. Then Milstein had a spark: myeloma cells could divide indefinitely, while B cells produced specific antibodies. What if the two could be fused?
The procedure sounded simple but was full of subtleties: immunize mice with emulsified sheep red blood cells, harvest spleen B cells that secreted specific antibodies, fuse them with antibody-deficient myeloma cells using polyethylene glycol (PEG), then culture the mixture in selective medium and watch for hybrid cells that both survived and produced antibodies.
▲ Principle of Hybridoma Technology
Weeks later, Köhler spotted living cells in one well secreting antibodies specific only to sheep red blood cells. “We did it!” he cried, rushing into Milstein’s office with trembling hands, clutching the assay results.
In August 1975, Nature published their two-page paper. Few realized it marked the dawn of a therapeutic revolution. Nine years later, the Nobel Prize in Physiology or Medicine went to Köhler, Milstein, and Jerne, who had proposed the theory of immune specificity. On stage, Milstein said with a smile: “We only meant to create a research tool. We never thought it would become a medicine that saves lives.”
▲ The two-page paper on hybridoma technology, published in Nature in 1975
The true gift was that the “tool” didn’t stay locked in a lab. After publication, requests for hybridoma cell lines poured in. Milstein’s group mailed frozen vials worldwide-without charging a penny. Many researchers later recalled finding handwritten notes tucked into the packages: “Mind the culture temperature. Write to us if you run into trouble.”
Lawmakers at the time criticized the MRC for “failing to protect intellectual property.” Looking back, that openness laid the foundation for the antibody industry. Back then, biotech was still in its infancy, and universities in the US and UK were not even allowed to patent government-funded discoveries. Köhler and Milstein only wanted to answer a basic question-how antibodies are produced-yet unwittingly sowed the seeds of hope for millions of patients.
In the decades since, antibody science has never dimmed. At least six researchers have won the Nobel or Lasker Awards for breakthroughs in antibody drug development-from Susumu Tonegawa, who uncovered the genetic basis of antibody diversity, to James Allison, pioneer of immune checkpoint therapy. Behind each award lies a legacy of Köhler and Milstein’s original exploration.
▶ Chapter One: The “Mouse Antibody Dilemma”
The hybridoma technology was like discovering a new continent. Soon, mouse monoclonal antibodies became lab superstars, helping researchers identify cancer and infection targets. But attempts to make them medicines ran into a wall.
In 1986, the first approved antibody drug, muromonab-CD3, was launched to prevent surgery rejection. Hopes were high-until clinical practice revealed harsh truths.
Patients’ immune systems recognized the mouse antibody as foreign, producing anti-drug antibodies (ADAs) that cleared it within days. Severe allergic reactions, even shock, occurred. Worse still, the drug’s half-life in humans was only 1-2 days, forcing frequent infusions.
A physician on the trial recalled: “One surgery patient lost all benefit within two weeks. The rejection episodes progressed despite treatment, highlighting the urgent need for better solutions.”
Thus emerged the “three sins” of mouse antibodies: strong immunogenicity, short half-life, and weak effector function. The only way forward was to humanize them.
The first breakthrough came with chimeric antibodies. In 1984, scientists grafted mouse variable regions onto human constant regions-67% human, 33% mouse. This reduced immunogenicity significantly.
In 1994, abciximab, the first chimeric antibody, was approved to prevent thrombosis after stent placement. It targeted platelet GPIIb/IIIa, cutting clotting risk by 30% with far fewer allergic reactions than muromonab.
But progress didn’t stop there. In 1986, Winter’s Cambridge team advanced to humanized antibodies: only the six antigen-recognizing loops (CDRs) were retained from mouse antibodies, making them ~90% human.
In 1997, daclizumab, the first humanized antibody, was approved for surgery rejection. ADA incidence fell below 10%, and half-life extended to 14 days. Finally, patients and physicians had relief.
Then, in 2002, came an important milestone: the first fully human antibody, adalimumab. Unlike its predecessors, it wasn’t engineered from mouse antibodies, but discovered directly from human antibody libraries using phage display.
▲ The first-generation ‘Blockbuster Drugs’-adalimumab (Humira)
Adalimumab, targeting TNF-α, transformed rheumatoid arthritis care and went on to treat nine autoimmune diseases. It became a groundbreaking treatment for multiple autoimmune diseases, benefiting millions of patients.
From mouse to chimeric, to humanized, to fully human-this 27-year journey was filled with gene splicing, protein expression, animal testing, and countless dead ends. Yet each step moved closer to safer, more effective therapies.
▶ Chapter Two: Diverging Paths-The Rise of ADCs and Bispecifics
As fully human antibodies transformed the treatment of autoimmune disease, researchers in oncology asked: could antibodies also deliver therapeutic compounds more effectively against tumors? This led to the development of antibody-drug conjugates (ADCs).
In an ADC, the antibody acts as a delivery vehicle, the active compound is the therapeutic payload, and the linker controls the release-ensuring that the medicine is delivered inside target cells.
The first ADC, gemtuzumab ozogamicin (Mylotarg), received approval in 2000 for acute myeloid leukemia (AML). It combined an anti-CD33 antibody with calicheamicin, a DNA-interacting compound. However, safety challenges led to its withdrawal in 2010, temporarily slowing progress in the field.
Later, researchers at Genentech improved the linker design, allowing release only within tumor lysosomes. This innovation enabled greater precision.
In 2013, trastuzumab emtansine (T-DM1) was approved for HER2-positive breast cancer. It used the microtubule inhibitor DM1 with an optimized drug-antibody ratio (DAR). Clinical studies showed improved outcomes with a more favorable safety profile compared with previous approaches.
In 2019, trastuzumab deruxtecan (T-DXd) marked another step forward. Its payload, a topoisomerase I inhibitor (DXd), had the ability to affect neighboring tumor cells, a phenomenon sometimes called the “bystander effect.” In 2022, it also gained approval for HER2-low breast cancer, expanding treatment options for patients who were not previously eligible for HER2-targeted therapies.
▲ The Application of Anti-HER2 Antibody Therapeutics in Oncology
At the same time, bispecific antibodies (bsAbs) were emerging. These molecules can recognize two targets simultaneously-for example, binding to a tumor-associated antigen and a T cell receptor, thereby facilitating closer interaction between immune cells and tumors.
In 2009, catumaxomab, directed against EpCAM and CD3, was approved for malignant ascites, though later withdrawn. A major milestone came in 2014 with blinatumomab, a CD19×CD3 BiTE, which achieved encouraging remission rates in acute lymphoblastic leukemia (ALL). Although it required continuous infusion due to a short half-life, its clinical impact was significant.
More recent designs have improved durability and activity. Glofitamab (2022) introduced a “2+1” binding format for enhanced target engagement, while zenocutuzumab (2023) focused on HER2 and HER3 in cancers with rare NRG1 fusions.
▲ Common Modes of Action and Antigen Targets of Bispecific Antibodies in Clinical Trials
ADCs and bispecific antibodies represent two major approaches in the advancement of antibody therapeutics-one centered on targeted delivery of therapeutic agents, the other on enhancing immune system engagement. Despite their different mechanisms, both aim to make cancer treatment more precise and patient-centered.
▶ Chapter Three: Clinical Echoes-From “No Options” to“ Living with Cancer”
In 2004, 28-year-old Lisa was diagnosed with HER2+ metastatic breast cancer. Doctors told her she likely had less than three years. She entered a trastuzumab trial. Three months later, scans showed her lung metastases had shrunk by 30%. A year later, they were nearly gone. She is 52 years old and still alive and well.
Her story symbolizes how trastuzumab reshaped outcomes. In 1998, HER2+ patients’ 5-year survival was just 40%. With trastuzumab, pertuzumab, T-DM1, and T-DXd, it now exceeds 85%.
In 2018, Mike, a patient with relapsed diffuse large B-cell lymphoma, had no options left until he entered a trial for polatuzumab vedotin, an anti-CD79b ADC. Six months later, he achieved complete remission. “Before, chemo left me too weak to hold a cup. Now I play basketball twice a week,” he said.
For autoimmune diseases, antibody drugs were equally transformative. Sarah, age 45, had rheumatoid arthritis so severe she couldn’t tie her shoes. After starting adalimumab in 2019, she regained function and confidence, no longer feeling like a burden on her family.
Behind every patient’s recovery are vast clinical trials and relentless innovations: Fc engineering for stronger ADCC, recycling antibodies via FcRn for longer half-lives, and subcutaneous formulations for convenience.
As Dr. Carter put it: “We don’t make antibodies for papers or prizes-we make them so patients can live normal lives.”
▶ Chapter Four: The Unsung Heroes-Biomanufacturing Revolution
Few realize that manufacturing is often the bottleneck between lab discovery and clinical medicine.
When muromonab-CD3 launched in 1986, producing 1 gram of antibody required tens of liters of culture, with purity issues so severe that entire batches were discarded.
“The transformation began with CHO cells (Chinese hamster ovary cells). These cells can efficiently express human proteins and are easy to culture. In the 1990s, scientists started genetically engineering CHO cells to improve antibody expression. By 2000, antibody yields reached 5-8 g/L; today, optimized CHO cells can produce over 10 g/L-equivalent to what previously required 10 liters of culture now achieved in just 1 liter.
▲ Chinese hamster ovary (CHO) cells-the most widely used cell line for antibody production today
Another revolution was the rise of single-use bioreactors. Unlike stainless steel tanks requiring lengthy sterilization, disposable systems cut contamination risks and improved scalability. Today, >70% of antibody drugs are made this way.
▲ Single-use bioreactor
Costs plummeted too-from over $10,000 per gram in the 1990s to just a few hundred dollars today-making antibody drugs far more accessible.
As one engineer said: “Without advances in manufacturing, even the best science couldn’t reach patients.”
▶ Chapter Five: The Next Fifty Years-AI and Beyond
In 2024, the Nobel Prize in Chemistry honored AlphaFold, the AI tool predicting protein structures with unprecedented accuracy. For antibody scientists, this was game-changing. Design could now be guided by computation, boosting efficiency tenfold.
Dreams once impossible are moving closer. Oral antibodies are being engineered, such as AI-designed single-domain antibodies (VHHs) resistant to digestion, which in 2025 showed success in monkeys against IL-23.
Brain-penetrating antibodies are also emerging. By linking one arm to TfR, they can cross the blood-brain barrier, opening new paths for Alzheimer’s and brain cancer therapy.
Even multi-specific antibodies are advancing-targeting PD-1, CTLA-4, and LAG-3 simultaneously to supercharge immune responses against resistant cancers.
▲ 50 Years of Trials and Triumphs in the World of Antibody Therapeutics
As Dr. Carter reflected: “Fifty years ago, we couldn’t even imagine a human antibody. Now we’re designing oral and brain-penetrant antibodies. In the next fifty years, we may see personalized antibodies tailored to each patient’s genome, or ‘smart antibodies’ that adapt to tumors in real time.”
▶ Epilogue: Fifty Years On, the Long March Continues
In June 2025, students of Köhler and Milstein gathered in Cambridge for a memorial. Among the displays were original frozen hybridoma vials labeled in handwriting: “Anti-sheep red blood cell, 1975.8.7.”
From lab “tool” to billion-dollar “powerful biological tools,” from mouse antibody pitfalls to the flowering of ADCs and bispecifics, from clunky steel bioreactors to high-yield disposable systems-the journey has been full of challenges and triumphs.
Behind every milestone were students pulling all-nighters for Western blots, doctors logging bedside data, engineers fine-tuning culture media, and patients bravely joining trials. They are all part of this long march.
Today’s young scientists continue the relay: observing antibody-cell binding under microscopes, simulating designs with AI, refining new manufacturing methods. They know the story is unfinished-and the next fifty years await new impossibilities to be broken.
As the saying goes: Medical progress is not the flash of one genius, but the relay of generations. The past fifty years of antibody therapeutics have proven this true. The next fifty will too.
▶ Acknowledgment
In writing this reflection on fifty years of antibody therapeutics, I drew inspiration from a landmark review: “Fifty years of monoclonals: the past, present and future of antibody therapeutics” (Nature Reviews Immunology, Aug 2025), authored by Genentech scientists Andrew C. Chan, Greg D. Martyn, and Paul J. Carter.
My hope was to translate complex technical evolution into a story for the general reader-a warm “scientific epic” tracing the arc of monoclonal antibodies: from the 1975 hybridoma breakthrough, through mouse-to-human antibody engineering, to ADCs, bispecifics, and CAR-T therapies that continually expand the frontiers of medicine, heralding the golden age of antibody therapeutics.
推荐阅读
三优十周年｜针对宠物疾病的单克隆抗体药物
三优十周年｜CAR-T细胞治疗现状挑战与未来
三优十周年｜服务篇-体外药效筛选解决方案
三优十周年｜AI-STAL篇-Anticalin靶向蛋白库
三优十周年｜靶点篇-自身免疫疾病靶点介绍
创新前沿｜三优生物类器官产学研联盟启航
中和抗体“硬核出击”，斩断CHIKV病毒感染链
三优十周年｜服务篇-靶点至PCC药物研发
三优十周年｜靶点篇-六大恶性肿瘤靶点介绍创新前沿｜2025ADC热门靶点与技术创新启示三优十周年｜AI-STAL篇-高等电点单域抗体库
6类分子形式之环状多肽分子产生系统解决方案
轻松玩转三优oneClick+线上9大云程序
6类分子形式之mRNA分子产生系统解决方案
三重好礼相送暨四种抗体制备解决方案全新上线
关于三优生物
三优生物是一家以“让天下没有难做的创新生物药”为使命，以超万亿分子库和智能科技驱动的生物医药高科技企业。
公司致力于打造全球顶尖的原创新药创新工场。公司以智能超万亿分子库（AI-STAL）为核心；以干湿结合、国际领先的创新生物药智能化及一体化研发平台为依托；以多样化的业务模式推动全球创新药物的研发及产业化。
公司总部位于中国上海，在亚洲、北美洲、欧洲等多地建立了业务中心，形成了全球化的业务网络，现有投产及布局的研发及GMP场地20000多平方米。
公司已与全球2000多家药企、生技公司等建立了良好的合作关系，已赋能1200多个新药研发项目；已完成50多个合作研发项目，其中10多个协同研发项目已推至IND及临床研发阶段。
公司已申请130多项发明专利，其中30多项发明专利已获得授权，并获得了国家级高新技术企业、上海市专精特新、ISO9001、ISO27001等10余项资质及体系认证。
