在 R/R FL 中，独立审查委员会 (IRC) 评估的 ELM-2 (N=128) 结果显示客观缓解率 (ORR) 为 80%，其中 73% 的患者获得完全缓解 (CR)。在完全缓解者中，中位缓解持续时间 (DoR) 为 25 个月（95% 置信区间 [CI]：20 个月至无法评估 [NE]）。 在 R/R DLBCL 中， 根据 IRC 评估，ELM-2（N=127）在未接受过 CAR-T 治疗的患者中的结果显示 ORR 为 52%，其中 31% 达到 CR。在完全缓解者中，中位 DoR 为 18 个月（95% CI：10 个月至 NE）。 CAR-T 治疗后出现进展的 ELM-1（N=60）患者结果显示，经 IRC 评估，ORR 为 48%，其中 32% 达到 CR。在有反应的患者（n=29）中，中位 DoR 为 15 个月（95% CI：3 个月至 NE）。
Cycle 1: Odronextamob administered with steroid prophylaxis and step-up doses of 0.7/4/20 mg.（The step-up regimen was consisted of 0.7/4/20 mg administered as 0.2 mg and 0.5 mg on C1D1 and C1D2, 2 mg each on C1D8 and C1D9, and 10 mg each on C1D15 and C1D16. ） Cycle 2 through 4: 160 mg odronextamab on days 1, 8, and 15 of each cycle. Maintenance therapy: 320 mg odronextamab every 2 weeks.
Epcoritamab给药方案：
Day 1 (step-up dose 1): 0.16 mg SC Day 8 (step-up dose 2): 0.8 mg SC Day 15 (first full dose): 48 mg SC (hospitalize patient for 24 hr during first 48-mg dose) Day 22: 48 mg SC
Cycles 2 and 3
Days 1, 8, 15, and 22: 48 mg SC
Cycles 4-9
Days 1 and 15: 48 mg SC
Cycle 10 and thereafter
48 mg SC on Day 1 q28 days until disease progression or unacceptable toxicity
我主要想探究的一个问题，抗体的亲和力所带来的临床剂量的差异是否造成的疗效也有显著不同，以上可见Odronextamab给药剂量要高于Epcoritamab。
单纯从三线复发难治follicular lymphoma疗效数据来看：
Odronextamab
Results: Among 128 patients evaluated, 95% completed Cycle 1, and 85% completed ≥4 cycles. At 20.1 months' efficacy follow-up, ORR was 80.0% and complete response rate was 73.4%. Median duration of complete response was 25.1 months. Median progression-free survival was 20.7 months, and median overall survival was not reached. Discontinuation of odronextamab due to adverse events (AEs) occurred in 16% of patients. The most common treatment-emergent AEs were CRS (56%; grade ≥3 1.7% [1/60] with 0.7/4/20 mg step-up), neutropenia (39%), and pyrexia (38%). （https://pubmed.ncbi.nlm.nih.gov/39147364/）
Findings: Between June 19, 2020, and April 21, 2023, 128 patients (median age 65 years [IQR 55-72]; 49 [38%] female and 79 [62%] male) were enrolled and treated in the pivotal cohort (median follow-up 17·4 months [IQR 9·1-20·9]). The overall response rate was 82·0% (105 of 128 patients; 95% CI 74·3-88·3), with a complete response rate of 62·5% (80 of 128; 95% CI 53·5-70·9). The most common grade 3-4 treatment-emergent adverse event was neutropenia in 32 (25%) of 128 patients. Grade 1-2 cytokine release syndrome was reported in 83 (65%) of 128 patients; grade 3 cytokine release syndrome was reported in two (2%). Immune effector cell-associated neurotoxicity syndrome was reported in eight (6%) of 128 patients (five [4%] grade 1; three [2%] grade 2). Between Oct 25, 2022, and Jan 8, 2024, 86 patients (median age 64 years [55-71]; 37 [43%] female and 49 [57%] male) were enrolled and treated in the cycle 1 optimisation cohort. The incidence of cytokine release syndrome was 49% (42 of 86 patients; eight [9%] grade 2; none of grade 3 or worse), with no reported immune effector cell-associated neurotoxicity syndrome. （https://pubmed.ncbi.nlm.nih.gov/38889737/）
通过以上数据来看，两者数据相差不多，CR有部分区别，ORR相当，也就意味着CD3抗体亲和力的不同所带来激活能力的不同，在一定程度上可以通过剂量调节来平衡药效与安全性的关系。对于血液瘤而言。
序列层面来看，两者并不一致。Odronextamab采用共同轻链的技术防止抗体轻重链错配，而为了防止重链错配形成Homedimer，其中一个Fc采用了H435R, Y436F突变，该组合突变可以降低其与Protein A结合能力，从而便于后期目标双特异抗体的纯化。Epcoritamab由Genmab研发，采用Duobody双抗技术平台构建，在两条重链的CH3分别引入K409R和F405L突变，体外通过Fab exchange形成双特异性抗体。
Odronextamab
(Heavy chain)
EVQLVESGGG LVQPGRSLRL SCVASGFTFN DYAMHWVRQA PGKGLEWVSV ISWNSDSIGY
ADSVKGRFTI SRDNAKNSLY LQMHSLRAED TALYYCAKDN HYGSGSYYYY QYGMDVWGQG
TTVTVSSAST KGPSVFPLAP CSRSTSESTA ALGCLVKDYF PEPVTVSWNS GALTSGVHTF
PAVLQSSGLY SLSSVVTVPS SSLGTKTYTC NVDHKPSNTK VDKRVESKYG PPCPPCPAPP
VAGPSVFLFP PKPKDTLMIS RTPEVTCVVV DVSQEDPEVQ FNWYVDGVEV HNAKTKPREE
QFNSTYRVVS VLTVLHQDWL NGKEYKCKVS NKGLPSSIEK TISKAKGQPR EPQVYTLPPS
QEEMTKNQVS LTCLVKGFYP SDIAVEWESN GQPENNYKTT PPVLDSDGSF FLYSRLTVDK
SRWQEGNVFS CSVMHEALHN HYTQKSLSLS LG
(Light chain)
EIVMTQSPAT LSVSPGERAT LSCRASQSVS SNLAWYQQKP GQAPRLLIYG ASTRATGIPA
RFSGSGSGTE FTLTISSLQS EDFAVYYCQH YINWPLTFGG GTKVEIKRTV AAPSVFIFPP
SDEQLKSGTA SVVCLLNNFY PREAKVQWKV DNALQSGNSQ ESVTEQDSKD STYSLSSTLT
LSKADYEKHK VYACEVTHQG LSSPVTKSFN RGEC
(heavy chain)
EVQLVESGGG LVQPGRSLRL SCAASGFTFD DYTMHWVRQA PGKGLEWVSG ISWNSGSIGY
ADSVKGRFTI SRDNAKKSLY LQMNSLRAED TALYYCAKDN SGYGHYYYGM DVWGQGTTVT
VASASTKGPS VFPLAPCSRS TSESTAALGC LVKDYFPEPV TVSWNSGALT SGVHTFPAVL
QSSGLYSLSS VVTVPSSSLG TKTYTCNVDH KPSNTKVDKR VESKYGPPCP PCPAPPVAGP
SVFLFPPKPK DTLMISRTPE VTCVVVDVSQ EDPEVQFNWY VDGVEVHNAK TKPREEQFNS
TYRVVSVLTV LHQDWLNGKE YKCKVSNKGL PSSIEKTISK AKGQPREPQV YTLPPSQEEM
TKNQVSLTCL VKGFYPSDIA VEWESNGQPE NNYKTTPPVL DSDGSFFLYS RLTVDKSRWQ
EGNVFSCSVM HEALHNRFTQ KSLSLSLG
Epcoritamab
>A chain_anti-CD3E
EVKLVESGGGLVQPGGSLRLSCAASGFTFNTYAMNWVRQAPGKGLEWVARIRSKYNNYAT
YYADSVKDRFTISRDDSKSSLYLQMNNLKTEDTAMYYCVRHGNFGNSYVSWFAYWGQGTL
VTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPA
VLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAP
EFEGGPSVFLFPPKPKDTLMISRTPEVTCVVVAVSHEDPEVKFNWYVDGVEVHNAKTKPR
EEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLP
PSREEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFLLYSKLTV
DKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG
>B chain_anti-CD20
EVQLVESGGGLVQPDRSLRLSCAASGFTFHDYAMHWVRQAPGKGLEWVSTISWNSGTIGY
ADSVKGRFTISRDNAKNSLYLQMNSLRAEDTALYYCAKDIQYGNYYYGMDVWGQGTTVTV
SSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQ
SSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPEFE
GGPSVFLFPPKPKDTLMISRTPEVTCVVVAVSHEDPEVKFNWYVDGVEVHNAKTKPREEQ
YNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSR
EEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVDKS
RWQQGNVFSCSVMHEALHNHYTQKSLSLSPG
>C chain_anti-CD3E
QAVVTQEPSFSVSPGGTVTLTCRSSTGAVTTSNYANWVQQTPGQAFRGLIGGTNKRAPGV
PARFSGSLIGDKAALTITGAQADDESIYFCALWYSNLWVFGGGTKLTVLGQPKAAPSVTL
FPPSSEELQANKATLVCLISDFYPGAVTVAWKADSSPVKAGVETTTPSKQSNNKYAASSY
LSLTPEQWKSHRSYSCQVTHEGSTVEKTVAPTECS
>D chain_anti-CD20
EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRLLIYDASNRATGIPA
RFSGSGSGTDFTLTISSLEPEDFAVYYCQQRSNWPITFGQGTRLEIKRTVAAPSVFIFPP
SDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLT
LSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC
搞清楚这个问题，我们接下来看结构价态的区别是否会影响药效。
其实对我来说，一直想要搞清楚的是罗氏的两个CD3/CD20双抗，一个是2
:1结构，名为Glofitamab，另一个是1:1结构，名为Mosunetuzumab，的疗效问题。
通过罗氏的宣传材料中以及临床开发策略可知，Glofitamab和Mosunetuzumab有一定不同，Glofitamab对于aNHL疗效更有优势。
通过临床数据：
Mosunetuzumab
Results: Two hundred thirty patients were enrolled. Doses up to 2.8 mg and 60 mg were assessed in groups A and B, respectively; maximum tolerated dose was not exceeded. In group B (n = 197), common adverse events (≥ 20% of patients) were neutropenia (28.4%), cytokine release syndrome (27.4%), hypophosphatemia (23.4%), fatigue (22.8%), and diarrhea (21.8%). Cytokine release syndrome was mostly low-grade (grade ≥ 3: 1.0%) and mainly confined to cycle 1. Across the doses investigated (group B), best overall response rates were 34.9% and 66.2% in patients with aggressive and indolent B-NHL, respectively, and complete response rates were 19.4% and 48.5%. Among patients with a complete response, the median duration of response was 22.8 months (95% CI, 7.6 to not estimable) and 20.4 (95% CI, 16 to not estimable) in patients with aggressive and indolent B-NHL, respectively. （https://pubmed.ncbi.nlm.nih.gov/34914545/）
Glofitamab
Results: As of May 18, 2021, 258 pts were enrolled in
the previously specified cohorts. Median age was 64.0 (range, 22‒86)
years, 62.0% were male, and the median number of prior therapies was 3
(range, 1‒12). A total of 183 (70.9%) pts had aggressive NHL (aNHL), and
75 (29.1%) had indolent NHL (iNHL). Of the pts with aNHL, 98 had
diffuse large B-cell lymphoma, 26 had mantle cell lymphoma, 31 had
transformed follicular lymphoma (FL), and 11 had Richter's
transformation. All pts with iNHL had Grade (Gr) 1‒3a FL. Response rates
are reported across all doses investigated. Highest responses were seen
with the RP2D (2.5/10/30mg) in pts with aNHL .
At the clinical cut-off date (CCOD), median duration of follow-up in pts
with aNHL was 13.4 (range: 0‒36) months. In efficacy-evaluable pts with
aNHL (n=175), the overall response rate (ORR) was 53.7% and the CR rate
was 39.4%. Median duration of CR had not yet been reached (95%
confidence interval [CI]: 21.4‒not estimable [NE], n=69);
72.5% of pts with a CR (50/69) were still in CR at the time of
analysis. Median DoR (CR plus partial response) was 29.4 months (95% CI:
6.0‒NE; responders, n=94). In pts with iNHL (n=75), ORR was 81.3% and
CR rate was 69.3%; median follow-up was 8.6 (range: 0‒37) months. Median
duration of CR had not yet been reached (95% CI: 10.5‒NE, n=52);
82.7% of pts with a CR (43/52) were still in CR at the time of the
analysis. Median DoR had not been reached (95% CI: 10.5‒NE; responders,
n=61). A total of 149/258 pts (57.8%) experienced a serious adverse
event (AE). CRS was the most prevalent AE, occurring in 152/258 pts
(58.9%). The majority of CRS events were mild: Gr 1-2, 139 (53.9%) pts;
Gr 3, 9 (3.5%) pts; Gr 4, 4 pts (1.6%). Four pts (1.6%) experienced a
glofitamab-related AE that led to withdrawal of the study drug.
Ninety-two (35.7%) pts experienced a neurological AE; the majority of
events were Gr 1 (56/258; 21.7%) or Gr 2 (33/258; 12.8%). Three pts
experienced a Gr 3 neurological AE (facial paralysis, syncope,
radiculopathy), which were considered unrelated to glofitamab treatment.
Immune effector cell-associated neurotoxicity syndrome (ICANS)-like
events related to glofitamab occurred in 9 pts (3.5%); all events were
Gr 1 or Gr 2, and all but one (Gr 1 tremor) resolved at CCOD. （https://doi.org/10.1182/blood-2021-146845）
这两者临床数据就有很大的差异性，Glofitamab数据为ORR (aNHL): 53.7%/CR 39.4%; ORR (iNHL): 81.3%/CR 69.3%，Mosunetuzumab数据为ORR (aNHL): 34.9%; ORR (iNHL): 66.2%。所以结构的不同会导致疗效的不同。
在我曾经研发的过程中二价相比于单价，是有一定的杀伤优势的。但我也曾经考虑过一个问题，就是杀伤是不是有一定阈值，只要到了一定阈值，杀到一定程度是不是杀伤强弱的影响就不那么有区别，这是我一直以来在内心深处的想法，现在来看还是有一定区别。
在三线复发/难治性弥漫性大b细胞淋巴瘤，Glofitamab相比于Mosunetuzumab也表现出一定优势。Mosunetuzumab的ORR: 42.0%; CR: 23.9%（DOI: 10.1182/bloodadvances.2022009260 ），Glofitamab的ORR为52%，CR为39%（DOI: 10.1056/NEJMoa2206913 ）。
Glofitamab在多个适应症中表现出一定优势。所以结构价态的设计对于CD3多抗是有一定影响的。
序列如下
Mosunetuzumab
>CD3E Heavy Chain
EVQLVQSGAEVKKPGASVKVSCKASGYTFTNYYIHWVRQAPGQGLEWIGWIYPGDGNTKY
NEKFKGRATLTADTSTSTAYLELSSLRSEDTAVYYCARDSYSNYYFDYWGQGTLVTVSSA
STKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSG
LYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGP
SVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYGS
TYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSREEM
TKNQVSLSCAVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLVSKLTVDKSRWQ
QGNVFSCSVMHEALHNHYTQKSLSLSPGK
>CD3E Light Chain
DIVMTQSPDSLAVSLGERATINCKSSQSLLNSRTRKNYLAWYQQKPGQPPKLLIYWASTR
ESGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCTQSFILRTFGQGTKVEIKRTVAAPSV
FIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSL
SSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC
>CD20 Heavy Chain
EVQLVESGGGLVQPGGSLRLSCAASGYTFTSYNMHWVRQAPGKGLEWVGAIYPGNGDTSY
NQKFKGRFTISVDKSKNTLYLQMNSLRAEDTAVYYCARVVYYSNSYWYFDVWGQGTLVTV
SSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQ
SSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELL
GGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQ
YGSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSR
EEMTKNQVSLWCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKS
RWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK
>CD20 Light Chain
DIQMTQSPSSLSASVGDRVTITCRASSSVSYMHWYQQKPGKAPKPLIYAPSNLASGVPSR
FSGSGSGTDFTLTISSLQPEDFATYYCQQWSFNPPTFGQGTKVEIKRTVAAPSVFIFPPS
DEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTL
SKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC
Glofitamab
>SUBUNIT_1
QVQLVQSGAEVKKPGSSVKVSCKASGYAFSYSWINWVRQAPGQGLEWMGRIFPGDGDTDY
NGKFKGRVTITADKSTSTAYMELSSLRSEDTAVYYCARNVFDGYWLVYWGQGTLVTVSSA
STKGPSVFPLAPSSKSTSGGTAALGCLVEDYFPEPVTVSWNSGALTSGVHTFPAVLQSSG
LYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDEKVEPKSCDGGGGSGGGGSQAVVTQE
PSLTVSPGGTVTLTCGSSTGAVTTSNYANWVQEKPGQAFRGLIGGTNKRAPGTPARFSGS
LLGGKAALTLSGAQPEDEAEYYCALWYSNLWVFGGGTKLTVLSSASTKGPSVFPLAPSSK
STSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSL
GTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMI
SRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDW
LNGKEYKCKVSNKALGAPIEKTISKAKGQPREPQVYTLPPCRDELTKNQVSLWCLVKGFY
PSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALH
NHYTQKSLSLSPGK
>SUBUNIT_2
QVQLVQSGAEVKKPGSSVKVSCKASGYAFSYSWINWVRQAPGQGLEWMGRIFPGDGDTDY
NGKFKGRVTITADKSTSTAYMELSSLRSEDTAVYYCARNVFDGYWLVYWGQGTLVTVSSA
STKGPSVFPLAPSSKSTSGGTAALGCLVEDYFPEPVTVSWNSGALTSGVHTFPAVLQSSG
LYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDEKVEPKSCDKTHTCPPCPAPEAAGGP
SVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNS
TYRVVSVLTVLHQDWLNGKEYKCKVSNKALGAPIEKTISKAKGQPREPQVCTLPPSRDEL
TKNQVSLSCAVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLVSKLTVDKSRWQ
QGNVFSCSVMHEALHNHYTQKSLSLSPGK
>SUBUNIT_3
DIVMTQTPLSLPVTPGEPASISCRSSKSLLHSNGITYLYWYLQKPGQSPQLLIYQMSNLV
SGVPDRFSGSGSGTDFTLKISRVEAEDVGVYYCAQNLELPYTFGGGTKVEIKRTVAAPSV
FIFPPSDRKLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSL
SSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC
>SUBUNIT_4
DIVMTQTPLSLPVTPGEPASISCRSSKSLLHSNGITYLYWYLQKPGQSPQLLIYQMSNLV
SGVPDRFSGSGSGTDFTLKISRVEAEDVGVYYCAQNLELPYTFGGGTKVEIKRTVAAPSV
FIFPPSDRKLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSL
SSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC
>SUBUNIT_5
EVQLLESGGGLVQPGGSLRLSCAASGFTFSTYAMNWVRQAPGKGLEWVSRIRSKYNNYAT
YYADSVKGRFTISRDDSKNTLYLQMNSLRAEDTAVYYCVRHGNFGNSYVSWFAYWGQGTL
VTVSSASVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQES
VTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC
罗氏如今正在开展1期临床试验中评估englumafusp alfa与CD20×CD3双特异性抗体格菲妥单抗（glofitamab）联用，治疗复发/难治性非霍奇金淋巴瘤（r/r NHL）患者。
Englumafusp alfa是罗氏一种双特异性抗体样融合蛋白，可同时靶向B细胞上的CD19，以及T细胞等免疫细胞上的4-1BB，从而引发强烈的共刺激信号，增强和延长细胞活性，增强抗肿瘤活性。Englumafusp alfa在T细胞受体信号存在且严格依赖CD19交联的情况下，通过4-1BB激动作用对T细胞提供强烈的共刺激，与CD20/CD3 T细胞衔接双特异性抗体联用，可以进一步增强后者的抗肿瘤活性。
初步研究数据表明，该联合疗法在侵袭性r/r
NHL（r/r
aNHL）患者中显示出良好的抗肿瘤活性，最佳客观缓解率（BoR）为67.0%，完全缓解（CR）率为57.0%；二线治疗r/r侵袭性NHL患者的BOR为77.0%，CR率为77.0%。此外，联合治疗具有与格菲妥单抗单药相当的安全性和细胞因子释放综合征（CRS）情况。
