免疫检查点抑制剂治疗胃肠胰神经内分泌肿瘤的进展.pdf

综述267.外科理论与实践2 0 2 3年第2 8 卷第3期免疫检查点抑制剂治疗胃肠胰神经内分泌肿瘤的进展韩序，王文权综述楼文晖，刘亮审校（复旦大学附属中山医院普外科-胰腺外科，上海200032)摘要 靶向免疫检查点的免疫治疗发展迅速，近年来在胃肠胰神经内分泌肿瘤（gastroenteropancreatic neuro-endocrine neoplasm,GEP-NEN)中初步探索，但能否带来临床获益尚无定论。本文系统梳理免疫检查点抑制剂(im-munecheckpoint inhibitor,ICI)单药或双药联合治疗CEP-NEN的临床试验现状及疗效。结果表明，ICI在GEP-NEN中仍未取得突破性进展，治疗复发或转移性CEP-NEN有一定的抗肿瘤活性和安全性，但总体客观缓解率（objectiveresponse rate,ORR)较低。ORR与肿瘤分化程度呈负相关，分化差的胃肠胰神经内分泌癌可能更易获得临床缓解。双药与单药相比，疾病控制率更高，但不良反应更严重。鉴于错配修复基因缺陷和微卫星高度不稳定极为罕见，肿瘤突变负荷高(10 muts/Mb)的病人能从ICI免疫治疗中改善生存。未来期望进一步探索ICI与化疗、放疗、抗血管生成药物等在CEP-NEN的联合应用，有可能提高其抗肿瘤疗效，起到“1+1 2 的效果。临床应根据病理分化分级、免疫标志物、病情进展程度、病人身体状态等综合评估ICI免疫治疗的获益人群。关键词：免疫检查点抑制剂；胃肠胰神经内分泌肿瘤；免疫治疗；临床获益中图分类号：R735文献标志码：C文章编号：10 0 7-9 6 10(2 0 2 3)0 3-0 2 6 7-0 6D0I:10.16139/j.1007-9610.2023.03.015Emerging developments in immune checkpoint inhibitor therapy for gastroenteropancreatic neuroendocrine neo-plasmHAN Xu,WANG Wenquan,LOU Wenhui,LIU LiangDepartment of General Surgery-Pancreatic Surgery,Zhongshan Hospital,Fudan University,Shanghai 200032,ChinaAbstract Immunotherapies targeting immune checkpoints have undergone rapid evolution,and have been prelimi-nary explored in treatment of gastroenteropancreatic neuroendocrine neoplasm(GEP-NEN)in recent years.However,theirpotential to deliver tangible clinical benefits remains uncertain.In this article,we systematically reviewed the current sta-tus and eficacy of clinical trials,which evaluated immune checkpoint inhibitor(ICI)as monotherapy or in dual-ICI therapyfor GEP-NEN.Despite lacking substantial breakthroughs in GEP-NEN treatment,ICI demonstrated some antitumor activityand safety in treating recurrent or metastatic GEP-NEN,albeit with a generally low objective response rate(ORR).TheORR of ICI in GEP-NEN treatment exhibited a negative correlation with tumor differentiation,suggesting that poorly diffe-rentiated gastroenteropancreatic neuroendocrine carcinoma(GEP-NEC)might achieve better clinical responses.Diseasecontrol rate of dual-ICI therapy was higher than that of monotherapy.However,dual-ICI also got more severe side effects.Given the rarity of mismatch repair gene defects and high microsatellite instability(dMMR/MSI-H)in GEP-NEN,patientswith high tumor mutational burden(TMB-H10 muts/Mb)could get potentially benefit from ICI therapy.In the future,it isexpected to further explore the synergistic combined application of ICI with chemotherapy,radiotherapy,and antiangio-genic drugs in GEP-NEN,which may enhance its antitumor efficacy.Clinically,the benefit groups of ICI immunotherapyshould be evaluated comprehensively according to pathological grading,immune markers,disease progression,and pa-tients physical condition.Key words:Immune checkpoint inhibitor;Gastroenteropancreatic neuroendocrine neoplasm;Immunotherapy;Clinicalbenefit胃肠胰神经内分泌肿瘤（gastroenteropancreatic neuroen-docrine neoplasm,GEP-NEN)是一组起源于胃肠胰神经内分泌细胞的具有显著异质性的肿瘤，包括高分化的胃肠胰神基金项目：国家自然科学基金（8 18 7 19 41,8 18 2 7 8 0 7,8 18 7 2 36 6）通信作者：刘亮，E-mail:经内分泌肿瘤（gastroenteropancreatic neuroendocrine tumor，GEP-NET)和低分化的胃肠胰神经内分泌癌（gastroentero-pancreatic neuroendocrine carcinoma,GEP-NEC)。近年来，随着免疫治疗在许多恶性实体瘤中的成功应用,其在GEP-NEN中的研究也逐步展开。A1?268J Surg Concepts Pract 2023,Vol.28,No.3免疫检查点是免疫系统中的抑制性受体和信号通路，可调节机体免疫激活的程度，避免损害正常组织。免疫检查点可被肿瘤细胞利用进行免疫逃逸。免疫检查点靶向疗法是通过共抑制或共刺激信号等相关途径调节T细胞活性来杀伤肿瘤细胞。近年来，免疫检查点抑制剂(im-munecheckpointinhibitor,ICI)包括程序性死亡受体（配体）1(programmed death 1/programmed death ligand 1,PD-1/PD-L1)抑制剂、细胞毒性T淋巴细胞相关抗原4（cytotoxicTlymphocyteassociatedprotein4,CTLA-4)抑制剂等免疫治疗在实体瘤中展现出广阔前景。在NEN中,ICI的循证医学证据主要集中在小细胞肺癌和梅克尔细胞癌(Merkelcellcarcinoma)2-31,两者肿瘤突变负荷普遍较高，肿瘤微环境存在较高的免疫原性4。鉴于GEP-NEN的肿瘤异质性和数量较少，现阶段缺乏针对GEP-NEN的高质量临床试验数据，ICI能否给GEP-NEN病人带来临床获益未有定论。本研究收集截至2 0 2 1年12 月正式发表的关于ICI治疗GEP-NEN的高质量临床试验，进行系统梳理(4-1(见表1)，探讨ICI治疗GEP-NEN的研究现状及疗效。1研究指标1.1ICI治疗GEP-NEN的总体客观缓解率在纳人的8 项临床试验中,ICI治疗GEP-NEN的客观缓解率(objoctive responserate,ORR）在3%40%之间，置信表1近期ICI治疗GEP-NEN的高质量临床试验汇总Tab1Summary of recent high-quality clinical trials of ICI in GEP-NEN treatmentCEPPrior Sys-Prior systemicPartici-PrimaryEnrollment dis-Follow-upAuthor(s)YearCountry Study typeNETDifferentiationtemictreat-treatmentICITargetpantsNETsiteease status(months)cases(%)ment(%)protocolLung,thymus,pan-Well-differenti-Progression afterPhase IILong-actingSpartali-Yaocreas,gastrointesti-86ated(n=95),prior systemicPD-1Median2021Globalclinical116100%growth inhibi-zumabet al.llnal tract,gallblad-(74.1%)Poorlydifferen-treatment,distant(CD279)value 13.4trialtors(29.3%)(PDR001)der,unknowntiated(n=21)metastasesModeratetoProgression afterNasopharynx,well-differenti-Phase IIprior systemicIpilimumabPatelesophagus,gastroin-8ated(n=2),CTLA-4Until2021USAclinical19100%Not reportedtreatment,no+et al.171testinal tract,cer-(42.1%)Poorly diferen-+PD-131.0trialavailable options,nivolumabvix,vulva,unknowntiated(n=11),Ki-67 all 20%Unknown(n=6)Moderate toLung,thymus,gas-Progression afterPhase IIwell-differenti-IpilimumabPateltrointestinal tract,15prior systemicCTLA-4Until2020USAclinical32ated(n=8),100%Sunitinib(7%)+et al.l0lcervix,prostate,un-(46.9%)treatment,no+PD-115.0trialPoorly differen-nivolumabknownavailableoptionstiated(n=4)Majority with dis-Lung,pancreas,tant metastasesPhase IIStrosberggastrointestinal83All well-diffe-Chemotherapy(106/107),pro-Pembroli-Median2020Globalclinical10797.2%PD-1et al.51tract,liver,ovaries,(77.6%)rentiated(65.9%)gression afterzumabvalue24.2trialunknownprior systemictreatmentModerate toChemotherapyPhase IILung,thymus,gas-well-differenti-Local advancedIpilimumabKlein10(86%EP orCTLA-4Until2020Australiaclinical29trointestinal tract,ated(n=26),89.7%stage or distant+et al.19(34.5%)CAPTEM pro-+PD-126.0trialprostate,unknownpoorly diferen-metastasesnivolumabtocol)tiated(n=3)Metastasis,dis-Well-differenti-Phase I bease progressionLuPancreas,gastroin-32ated(n=8),ToripalimabUntil2020Chinaclinical40100%PRRT(21%)after prior sys-PD-1et al.181testinal tract,others(80.0%)poorly-differen-(JS 001)24.0trialtemictreatment,tiated(n=32)Ki-67 20%Local advancedstage or distantPhase I bLung,pancreas,Mehnert16All well-diffe-Everolimusmetastases,pro-Pembroli-Until2020Globalclinical41gastrointestinal70.7%PD-1et al.l1o0(39.0%)rentiated(7%)gression afterzumab24.0trialtract,othersprior systemictreatmentLocal advancedstage or distantVijayver-Phase IIThymus,pancreas,24Everolimusmetastases,pro-Pembroli-gia2020USAclinical29gastrointestinalC3100%PD-1Until 36.0(82.8%)(31.7%)gression afterzumabet al.l4trialtract,kidneyprior systemictreatment,C3FP269.外科理论与实践2 0 2 3年第2 8 卷第3期区间大。ORR最低的是Yao等报道的Spartalizumab临床研究，仅为3%(9 5%CI:0.010.10)。Sp a r t a l i z u ma b 临床研究人组8 6 例病例数最多；类似的，KEYNOTE-158临床研究5)纳入8 3例，0 RR也仅为5%（9 5%CI:0.010.12）。相比之下，ORR最高的是CA209-538临床研究9,竞达到40%（9 5%CI：0.120.74）。以上研究表明，单独应用ICI治疗GEP-NEN疗效有限，多数临床研究ORR均 10%。经过荟萃分析森林图整合，合并的总0 RR仅为5%（9 5%CI:0.030.08）（见图1），与入组病例数多的Spartalizumab及KEYNOTE-158临床研究结果基本吻合。此外，森林图显示纳人的文献异质性较大，这也解释了ORR置信区间大的原因1.2ICI治疗分化好GEP-NET、分化差GEP-NEC亚组分析在这些临床试验中，分化好（well-differentiated，WD）GEP-NET及分化差(poorlydifferentiated,PD)GEP-NEC的例数分别为18 0 例及6 9 例。森林图亚组分析显示ICI治疗分化好GEP-NET的总体ORR为4%(9 5%CI:0.010.07)，而分化差GEP-NET的总体ORR为2 2%(9 5%CI:0.060.39)，两组差异有统计学意义（X=15.5,P0.001)（见图2）。这说明ICI治疗GEP-NEN的ORR与肿瘤分化程度呈负相关,PD的GEP-NEC可能更易获得临床缓解。但其亚组病例数较少，GEP-NENWeightWeightStudyEventsTotalProportion95%cl(common)(random)Yao-20213860.030.01;0.1044.9%44.9%Patel-2021380.38 0.09;0.760.6%0.6%Patel-20202150.13 0.02,0.402.3%2.3%Strosberg-20204830.05 0.01;0.1231.8%31.8%Klein-20204100.40 0.12;0.740.7%0.7%Lu-20205320.16 0.05;0.334.3%4.3%Mehnert-20201160.060.00;0.304.8%4.8%Vijayvergia-20201240.04 0.00;0.2110.6%10.6%Commoneffectmodel2740.050.03;0.08100.0%Randomeffectsmodel0.050.03;0.08100.0%Heterogeneity./=46%,t20.0001,p=0.070.10.20.30.40.50.60.7图1森林图显示ICI治疗GEP-NEN的ORRFig1Forest plot depicting ORR in GEP-NEN treated with ICIWDGEP-NETWeightWeightStudyEventsTotalProportion95%cl(common)(random)Ya0-20212650.030.00;0.1146.6%46.6%Patel-2020070.000.00;0.413.5%3.5%Strosberg-20204830.050.01;0.1242.6%42.6%Klein-2020020.000.00;0.840.6%0.6%Lu-2020170.140.00;0.581.4%1.4%Mehnert-20201160.060.00;0.305.4%5.4%Commoneffectmodel1800.040.01;0.07100.0%Randomeffectsmodel0.040.01;0.07100.0%Heterogeneity.=0%,t2=0,p=0.9500.20.40.60.8PDGEP-NECWeightWeightStudyEventsTotalProportion95%cl(common)(random)Ya0-20211210.050.00;0.2461.4%30.4%Patel-2021380.380.09;0.764.5%14.2%一-Patel-2020280.250.03;0.655.7%16.0%Klein-2020470.570.18;0.903.8%12.7%Lu-20204250.160.05;0.3624.7%26.7%Commoneffectmodel690.12 0.05;0.19100.0%Randomeffectsmodel0.220.06;0.39100.0%Heterogeneity./2=65%,t2=0.0215,p=0.02厂0.20.40.60.8A comparative study with significant statistical difference(4%us.22%,x*=15.5,P0.001).图2森林图亚组分析显示ICI治疗分化好GEP-NET及分化差GEP-NEC的ORRFig22 Subgroup analysis via forest plot illustrating ORR of well-differentiated GEP-NET and poorly differentiated GEP.NEC treated withICI?270J Surg Concepts Pract 2023,Vol.28,No.3异质性较大，仍需更多的临床试验数据进一步证明1.3单药ICI或双药ICI联合治疗GEP-NEN的亚组分析全组中仅SWOGS1609DARTl6-7及CA209-538/9临床试验是CTLA-4联合PD-1双药ICI治疗,其他临床试验均为单药PD-1治疗GEP-NEN。森林图亚组分析显示，单药ICI治疗GEP-NEN的ORR为5%(9 5%CI:0.020.07)CTLA-4联合PD-1双药ICI治疗GEP-NEN的ORR为2 3%（9 5%CI：0.090.36），两组差异有统计学意义（X=17.3,P0.001)(见图3)。双药ICI联合阻断CTLA-4和PD-1/PD-L1能发挥协同作用,重新激活受抑制的效应T细胞，进而显著增强肿瘤细胞免疫。1.4ICI治疗GEP-NEN的生存分析在这些临床试验中，人组标准基本为局部晚期或远处转移的病人。由于人组病人原发灶的同质性较差，因此，中位总生存期(median overall survival,mOS)在各项临床研究中差异较大。中位无进展生存期（median progression-freesurvival，mPF S）在2.2 4.8 个月，SWOG S1609DARTl6l和CA209-53819两项ICI双药临床研究中mPFS分别为4.0 个月和4.8 个月（见表2），双药ICI联合阻断CTLA-4和PD-1/PD-L1相较单药ICI可能会显示出较多的生存获益，但SWOGS1609DART研究中50%的病人存在3/4级的严重不良反应,31.5%的病人因此终止治疗。针对高度异质性的GEP-NEN,未来仍需进行同质性大样本的临床试验以明确双药ICI免疫治疗的疗效和不良反应。2 项临床研究表明PD-L1表达水平与治疗反应无明显相关性4.8,而我国NCT03167853临床研究显示肿瘤突变负荷9.9 muts/Mb的病人接受ICI能显著延长PFSI82讨论近年来ICI免疫治疗开启了肿瘤治疗的新纪元，但在GEP-NEN病人中还未取得突破性进展，,存在诸多局限。现阶段的临床试验表明，ICI治疗复发或转移性GEP-NEN有一定的抗肿瘤活性和安全性，但总体有效率低。单独应用ICI治疗GEP-NEN的疗效十分有限。另外笔者发现,ICI治疗GEP-NEN的ORR与肿瘤分化程度呈负相关，分化差的GEP-NEC可能更易获得临床缓解。但临床数据较少，仍需更多的研究进一步阐明对免疫治疗不敏感的CEP-NEN相关耐药机制。这也为提高ORR带来新的探索目标。低分化GEP-NEC具有很高的侵袭和转移能力。生物学特征与小细胞肺癌类似,常出现TP53、RB1、P16 基因突变。这些基因的异常使低分化GEP-NEC具有高度恶性的生物学行为，增殖迅速，肿瘤细胞突变率高，新生抗原易快速出现进而表达更高的免疫原性12-131;相反，ATRX-DAXX和MEN-1等基因改变在高分化而生长缓慢的GEP-NET病人中更常见4。同时,GEP-NET肿瘤较为惰PD-1monotherapyWeightWeightStudyEventsTotalProportion95%cl(common)(random)Ya0-20213860.030.01;0.1046.6%46.6%Strosberg-20204830.050.01;0.1233.0%33.0%Lu-20205320.160.05;0.334.4%4.4%Mehnert-20201160.060.00;0.305.0%5.0%Vijayvergia-20201240.040.00;0.2111.0%11.0%Commoneffectmodel2410.050.02;0.07100.0%Randomeffectsmodel0.050.02;0.07100.0%Heterogeneity.P=0%,t?0.0001,p=0.500.050.10.150.20.250.3CombinationtherapyofCTLA-4andPD-1WeightWeightStudyEventsTotalProportion95%cl(common)(random)Patel-2021380.38 0.09;0.7616.6%23.6%Patel-20202150.13 0.02;0.4063.1%49.3%Klein-20204100.40 0.12,0.7420.3%27.1%Commoneffectmodel330.23 0.09;0.36100.0%Randomeffectsmodel0.260.07;0.46100.0%Heterogeneity.P=36%,t2=0.0122,p=0.210.10.20.30.40.50.60.7A comparative study with significant statistical difference(5%us.23%,x2=17.3,P0.001).图3森林图亚组分析显示PD-1单药ICI及CTLA-4联合PD-1双药ICI治疗GEP-NEN的ORRFig 3 Subgroup analysis via forest plot demonstrating ORR of GEP-NEN treated with either monotherapy of PD-1 ICI orcombination therapy of CTLA-4 and PD-1 ICI271外科理论与实践2 0 2 3年第2 8 卷第3期表2近期免疫检查点抑制剂治疗GEP-NEN临床试验中的生存分析Tab 2Survival outcomes from recent clinical trials of ICI in the treatment of gastroenteropancreatic neuroendocrine tumorsNumber ofAuthorYearCountryStudy typeICIMedian OS(months)MedianPFS(months)casesNot reached(NET).3.8(NET),Yao et al.ll2021GlobalPhaseIl clinical trial116Spartalizumab(PDRO01)6.8(GEP-NEC)1.8(GEP-NEC)Patel et al.1712021USAPhase Il clinical trial19Ipilimumab+nivolumab8.92.0Patel et al.loi2020USAPhaseII clinical trial32Ipilimumab+nivolumab11.04.0Strosberg et al./52020ClobalPhaseIl clinical trial107Pembrolizumab24.24.1Klein et al.,912020AustraliaPhase II clinical trial29Ipilimumab+nivolumab14.784.829.1(PD-L110%),3.8(PD-L110%),Lu et al.s82020ChinaPhaseIbclinical trial40Toripalimab(JS 001)7.2(PD-L110%)2.2(PD-L110%)Mehnert et al.fiol2020GlobalPhaseI bclinical trial41Pembrolizumab21.0(pNET)4.5(pNET)Vijayvergia et l.4l2020USAPhase Il clinical trial29Pembrolizumab5.12.2pNET:pancreatic NET性，肿瘤突变负荷明显低于CEP-NEC。因此,如何将CEP-NEN从弱免疫原性的“冷肿瘤”转变为具有高突变负荷及新生抗原、富含肿瘤浸润淋巴细胞（tumorinfiltratinglympho-cytes,TIL)的 热肿瘤 可能是未来关注的方向!13。PD-1主要在活化效应CD8*T细胞上表达，以单体的形式存在于细胞表面。PD-L1是PD-1的主要配体,PD-L1在肿瘤细胞和其他免疫细胞中表达，在免疫应答的负性调控方面发挥重要作用。临床研究发现，ICI阻断PD-1/PD-L1信号通路促进肿瘤抗原特异性CD8T细胞的活化和增殖，增强免疫反应，从而发挥杀伤肿瘤细胞的作用。初始T细胞表面不表达CTLA-4,在T细胞活化后开始表达。CTLA-4的配体CD80/CD86主要表达在抗原提呈细胞(antigen-present-ingcell,APC),CTLA-4与其结合抑制T细胞活化发挥负向调节免疫应答作用。此外，调节性T细胞（regulatorycell，Treg）一方面与T细胞上的CD28竞争性结合APC上的CD80/CD86位点，进而抑制T细胞激活；另一方面，通过CTLA-4介导Treg细胞对APC表面CD80/CD86反式内吞,降低其表达来减少对T细胞表面CD28的激活15。此外，Treg还可增加PD-1和效应T细胞的游离PD-L1。综上所述，Treg对T细胞产生多重抑制作用。因此,ICI联合阻断CTLA-4和PD-1/PD-L1会协同拮抗肿瘤微环境中Treg介导的免疫抑制，进而显著增强肿瘤细胞免疫6。本文的结果与此类似。双药ICI治疗GEP-NEN,与单药相比，疾病控制率较高，但双药治疗不良反应较严重，病人的耐受性较差。临床应根据病理分化分级、免疫标志物、病情进展程度、病人身体状况等综合评估ICI免疫治疗的获益人群。本研究表明,PD-L1表达水平与治疗反应无明显相关性，可能是肿瘤异质性的原因。根据肿瘤原发部位不同，PD-LI表达差异较大2。肿瘤突变负荷高的病人接受ICI显著延长PFS。且由于错配修复基因缺陷和微卫星高度不稳定在GEP-NEN病人中极其罕见，因此，错配修复基因/微卫星稳定性的免疫标志物价值十分有限17 1。研究表明TIL阳性浸润的GEP-NEC病人生存期较短2 1。因此，笔者推测肿瘤突变负荷高（10 muts/Mb）以及TIL阳性浸润的病人可能从ICI免疫治疗中获益。最新研究揭示GEP-NEN的免疫微环境，将转移样原发型（metastasis-likeprimary，M L P)-1亚型鉴定为具有免疫相关基因的广泛而稳定激活的免疫高表型。发现MLP-1亚型存在高表达PD-L1、PD-L 2、L A G 3、ID O-1和C10orf54。为未来GEP-NEN中精确ICI免疫治疗提供了依据8。3结语目前,ICI免疫治疗的研究仍面临着巨大挑战。如何让ICI与化疗、放疗、抗血管生成药物联合应用并改善病人生存；如何将GEP-NEN从弱免疫原性的“冷肿瘤 转变为具有高突变负荷及新生抗原、富含TIL的“热肿瘤”；如何有效控制ICI免疫治疗相关不良事件的发生；如何精准定位ICI免疫治疗的受益人群，以上这些问题均需进一步关注和研究。参考文献川 POLEEIN,HERMANS B C M,VANDER ZWANJM,et al.Long-term survival in patients with gastroenteropan-creatic neuroendocrine neoplasms:a population-basedstudyJ.Eur J Cancer,2022,172:252-263.2 OTT P A,ELEZ E,HIRET S,et al.Pembrolizumab in pa-tients with extensive-stage small-cell lung cancer:resultsfrom the phase I b KEYNOTE-028 studyJ.J Clin On-col,2017,35(34):3823-3829.3 WEBER M M,FOTTNER C.Immune checkpoint inhibi-tors in the treatment of patients with neuroendocrine neo-plasiaJl.Oncol Res Treat,2018.41(5):306-312.4VIJAYVERGIA N,DASARI A,DENG M,et al.Pembro-lizumab monotherapy in patients with previously treatedmetastatic high-grade neuroendocrine neoplasms:jointanalysis of two prospective,non-randomised trialsJ.Br JCancer,2020,122(9):1309-1314.（本文编辑：许华芳）272J Surg Concepts Pract 2023,Vol.28,No.35STROSBERG J,MIZUNO N,DOI T,et al.Efficacy andsafety of pembrolizumab in previously treated advancedneuroendocrine tumors:results from the phase IIKEYNOTE-158 studylJ).Clin Cancer Res,2020,26(9):2124-2130.6PATEL S P,OTHUS M,CHAE Y K,et al.A phase IIbasket trial of dual anti-CTLA-4 and anti-PD-1 blockadein rare tumors(DART SWOG 1609)in patients with non-pancreatic neuroendocrine tumorsJ.Clin Cancer Res,2020,26(10):2290-2296.7PATEL S P,MAYERSON E,CHAE Y K,et al.A phaseI basket trial of dual anti-CTLA-4 and anti-PD-1 block-ade in rare tumors(DART)SWOG S1609:high-gradeneuroendocrine neoplasm cohortJJ.Cancer,2021,127(17):3194-3201.8 LU M,ZHANG P,ZHANG Y,et al.Efficacy,safety,andbiomarkers of toripalimab in patients with recurrent ormetastatic neuroendocrine neoplasms:a multiple-centerphase I b trialJJ.Clin Cancer Res,2020,26(10):2337-2345.9KLEIN O,KEE D,MARKMAN B,et al.Immunot