← 返回首页

ADC/双抗license-out CMC data room红旗:海外药企尽调最容易抓出的20个质量问题

海外buyer对中国ADC、双抗license-out交易的CMC尽调清单:cell line genetic stability、DAR分布、linker稳定性、comparability证据、impurity控制、stability program与data integrity——20条最容易被抓出的红旗与中国出让方的补救方式。

陈然
陈然最后更新:

本文只解决什么 / 不解决什么

这篇文章只解决一个非常具体的问题:你是中国创新药企(biotech或pharma)打算把一款ADC、双特异性抗体或多特异性产品license-out给一家美国/欧盟MNC(默沙东、阿斯利康、罗氏、辉瑞、艾伯维等),对方的CMC due diligence team要进你的data room。你需要预判他们会抓哪些红旗、提前补哪些证据、什么程度的不完美还在可接受范围、什么样的unfinished work会让deal value打折。

不解决:license-out整体deal structure(见站内创新药license-out交易指南ADC license-out全球商业化双抗全球开发与license-out)、临床数据/PK/PD的尽调、IP尽调。本文只覆盖CMC dataroom——也就是buyer的technical DD team会真正翻阅的内容。

为什么2025-2026年这个问题特别紧迫

按ApexOnco和Cooley的行业统计,中国biotech的ADC和bispecific license-out在2024-2025年达到历史峰值——仅2025年一年中国出让的ADC交易合计授权款超过300亿美元(ROC、Hansoh、Kelun-Biotech、MediLink、Innovent、Akeso等),双特异性ADC(bsADC)成为2026年最热的细分。但close的deal只是表象——许多deal在LOI签署后的CMC DD阶段被cut value、rescope、或彻底killed。

按BioSlice和Premier Research的访谈,中国ADC/双抗在DD中常被抓出的问题集中在以下几类:

  • 细胞株genetic stability:30-60-90代稳定性数据缺失或不完整
  • DAR分布:非site-specific conjugation的batch-to-batch DAR variability
  • linker稳定性:plasma stability数据缺失或在human matrix里没做
  • comparability:从Phase 1/2 lots到planned Phase 3 / commercial process的comparability证据严重不足
  • 杂质控制:尤其是free payload / drug-related related impurity(FDRIs)的acceptance criteria + 稳定性trend
  • stability program:accelerated和real-time stability batch数量不够、container closure未确认
  • data integrity:实验记录、电子数据系统、签字/审计trail是否符合FDA/EMA标准
  • supplier qualification:mAb intermediate / drug-linker / DP fill-finish多个站点的qualification状态
  • analytical method validation:method transfer到MNC commercial site的可行性
  • regulatory commitments:之前在NMPA、FDA INDsubmissions中的agreement / commitments是否兑现

中国出让方的典型痛点:当MNC的DD team列出20-30个CMC concern后,sponsor往往两个月内堆数据回应——但data room里其实没有可呈现的体系。这种情况下deal valuation下降5-30%是常见结果。

CMC dataroom的标准结构(buyer期望)

MNC的DD team进入data room时希望看到一个按ICH M4 Module 3 + ICH Q5/Q6/Q11 framework组织的目录。中国biotech常见的dataroom问题是:按内部SharePoint习惯堆放("old data 2022"、"latest as of 2024-Q3"、"raw data backup"),DD team找一份stability report要花2小时——这本身就是红旗。

可被accept的dataroom结构:

/CMC_DataRoom/
  ├── 01_General_Information/
  │   ├── Product_Profile.pdf
  │   ├── Regulatory_History_Timeline.pdf
  │   └── List_of_Manufacturing_Sites_with_GMP_Certificates.pdf
  ├── 02_Drug_Substance/
  │   ├── 2.1_mAb_Intermediate/                # antibody portion
  │   ├── 2.2_Drug_Linker/                     # small molecule + linker
  │   ├── 2.3_Conjugated_DS/                   # final ADC / bispecific drug substance
  │   │   ├── 2.3.1_Manufacturer_and_GMP/
  │   │   ├── 2.3.2_Process_Description/
  │   │   ├── 2.3.3_Process_Controls_and_Validation/
  │   │   ├── 2.3.4_Specifications/
  │   │   ├── 2.3.5_Analytical_Methods_and_Validation/
  │   │   ├── 2.3.6_Reference_Standards/
  │   │   ├── 2.3.7_Container_Closure/
  │   │   └── 2.3.8_Stability/
  │   └── 2.4_Cell_Line_and_Cell_Banking/
  ├── 03_Drug_Product/
  │   ├── (same Module 3.2.P structure)
  │   └── ...
  ├── 04_Quality_Control_System/
  │   ├── Site_Master_Files.pdf
  │   ├── Quality_Manual.pdf
  │   ├── Annual_Product_Review.pdf
  │   ├── Deviation_Trending.pdf
  │   ├── CAPA_Effectiveness.pdf
  │   ├── Complaint_Log.pdf
  │   └── Recall_History.pdf
  ├── 05_Comparability/
  │   ├── Process_A_to_Process_B_Comparability.pdf
  │   ├── Lot_to_Lot_Comparability_Studies.pdf
  │   └── Site_to_Site_Comparability (if applicable).pdf
  ├── 06_Regulatory/
  │   ├── 6.1_NMPA/  ├── 6.2_FDA_IND/  ├── 6.3_EMA/  └── 6.4_Other/
  │   └── (each with submissions, agency responses, commitments)
  ├── 07_Audit_Reports/
  │   ├── Customer_Audits.pdf
  │   ├── 3rd_Party_Audits.pdf
  │   ├── NMPA_Inspection.pdf
  │   ├── FDA_PAI_or_Pre-License_Inspection (if any).pdf
  │   └── EMA_GMP (if any).pdf
  └── 08_Supply_Chain/
      ├── Critical_Raw_Material_List.pdf
      ├── Single_Source_Supplier_Risk.pdf
      └── Tech_Transfer_Documentation.pdf

DD team通常会有一个pre-prepared CMC questionnaire(200-400个问题),90%的问题可以从dataroom直接找到答案——剩下10%是refinement类问题。如果dataroom只是把SharePoint镜像过去,就需要sponsor team每天回应20-30个查询,DD周期被拖成6-9个月,buyer的兴趣窗口流失。

红旗1:细胞株genetic stability数据缺失

ADC的mAb intermediate和双抗的整个DS都依赖于细胞株稳定性。CHO(包括CHO-K1、CHO-GS、CHO-DG44)是行业标准,但每个公司的clone都需要按ICH Q5B做genetic stability evaluation。MNC期待看到的:

期望中国常见缺口
Master Cell Bank(MCB)建立报告clone selection rationale + 关键attribute测试clone selection rationale不完整
Working Cell Bank(WCB)qualificationidentity、purity、安全性测试adventitious agent panel不全(特别是bovine viral panel)
Genetic stability studyLIVCD(limit of in vitro cell age for production)n+10 generations30-60-90代数据缺失或只做到MCB
Karyotype analysisn=30 cells minimum at MCB and post-LIVCDn=20以下
Sequence integrity全序列测序at MCB + post-LIVCD只做了短segment测序
Productivity stabilitytiter、product quality(aggregates、glycan、charge variants)at multiple cell ages只有titer没有quality attributes

补救路径:如果MCB建立超过3年了但没做完整的genetic stability,sponsor可以做"retrospective characterization"——用现有stored cells做30/60/90代培养、做end-of-production cell的characterization。这通常花3-4个月、几十万美元,但比buyer发现gap后大幅压价好很多。

红旗2:DAR分布的batch-to-batch variability

ADC的DAR(drug-to-antibody ratio)是critical quality attribute。MNC会看:

期望红旗
DAR测定方法UV/Vis + HIC-HPLC + LC-MS(至少两种正交方法)只用UV/Vis
DAR均值batch-to-batch variation typically < ±0.3 around target(如target 4.0 → range 3.7-4.3)variation > ±0.5
DAR分布DAR0/DAR2/DAR4/DAR6/DAR8 species的mass distribution稳定不同批次species分布飘移明显
Free payload的批间变化typically 0.5-2% target个别批次超出一倍
Site-specific conjugation的homogeneityfor site-specific:DAR2 species > 80%,DAR0 + DAR4 each < 10%DAR distribution wider than expected

双抗的同类问题:mismatched light chain pairing、HC-LC ratio、knob-into-hole assembly purity——这些是bispecific的"DAR equivalent"。期待values:correctly paired species > 90-95%(commercial process应稳定)。

补救路径:如果batch-to-batch variation高,sponsor可以提供"扩展batch dataset"(10-20批的trend analysis)+ "process robustness study"(critical process parameters in DOE范围内的variation)来证明characterization scope。如果variation truly太大,需要承认process仍在optimization期、并展示一份process improvement plan。

红旗3:linker稳定性数据不在human plasma中

linker stability测试有四个level,每个level都对ADC的safety/efficacy判断不同:

Level测试体系说明
1PBS或simple buffer体外稳定性,最低要求
2mouse / rat plasmapreclinical
3monkey plasmapreclinical, more clinically relevant
4human plasma临床相关;MNC一定会要看

中国厂常见缺口:linker stability只在rodent plasma里做了;human plasma stability是post-Phase 1做的、且常常用surrogate(如pooled normal human plasma而不是patient population的plasma)。MNC的DD会问:

  • Linker在human plasma里的half-life是多少?
  • 如果是cleavable linker,cleavage products是什么?bioactive metabolite的identity是否characterized?
  • 如果是non-cleavable,full ADC catabolism release的active species是什么?
  • 是否做过with antibody binding vs free linker的稳定性比较?

补救路径:linker stability的human plasma study通常2-3个月可补上,CRO(如Sterling、Charles River、AAPharma)有相应能力。所有release/stability/characterization assays应当cross-reference这些linker稳定性数据。

红旗4:comparability证据不足

Process A(early phase clinical lots)到Process B(Phase 3 / commercial)的comparability是ADC/双抗最大的CMC风险。FDA和EMA的expectation按ICH Q5E:

期望中国常见缺口
Comparability protocol在做process change前pre-defined改完process后才做"retrospective comparability"
Lot数量至少3 PPQ lots from new process + 同样数量的reference lots只有1-2批新工艺的lot
Release attribute比较全部CQA一一对照只比较了部分attributes
Stability比较至少6个月accelerated + 12个月real-time(commercial process)只有3个月数据
Biological characterizationbinding、potency、target engagement、higher-order structure只测了potency assay
Statistical analysis至少两个batch的pairwise comparison + tolerance interval只有means的简单对比

双抗的同类问题:从transient或stable pool process转到commercial CHO line的comparability、cell line从ATUM或GenScript转到internal的comparability、process scale-up(200L → 2000L → 12000L)的comparability。

补救路径:comparability需要时间——典型6-9个月。如果comparability不充分,MNC的DD team会要求sponsor提供:"comparability gap analysis + remediation plan"——sponsor明确承认哪些attributes还没比、什么时间能补、补不上时的regulatory risk。这比掩盖更有价值——MNC reviewer是professional,他们看出gap比sponsor自己列出来后果更严重。

红旗5:杂质控制——FDRIs和related impurities

ADC的杂质控制比一般biologic和小分子复杂。MNC的DD会问:

杂质类型期望红旗
Free drug / free payloadtypically NMT 1-2% at release,stability tracked没有定量方法
Free linkertypically NMT 1%没在specification里
Drug-related related impurities(FDRIs)identified + structures characterized + acceptance criteria"unspecified impurity NMT 0.5%" without identification
Antibody-related impuritiesaggregates、HMW、LMW、charge variants、glycan profilecharge variants drift not characterized
Linker-related impuritieshydrolysis products、reduction byproducts不列在specification
Process-related impuritiesresidual host cell protein、host cell DNA、protein A leached都按industry standard tested
Conjugation byproductsunconjugated mAb、over-conjugated speciesDAR0/DAR>target species的acceptance criteria缺失

双抗特有杂质:mismatched light chain、homodimers(HC-A/HC-A or HC-B/HC-B)、incorrect knob-into-hole pairing。这些应当用specific analytical methods(如mass spec或specific ELISA)quantify。

补救路径:杂质方法不全是个dealbreaker。但FDRIs identification(结构)必须在phase 3前完成——这是FDA和EMA都要求的。中国厂建议在phase 1-2过渡期就做完整的impurity characterization study。

红旗6:stability program的batch数量和时间

期望(commercial expectation)红旗
Drug Substance stability batches至少3个consecutive PPQ lots,6个月accelerated + 12个月real-time minimum at filing只有2批
Drug Product stability至少3个consecutive PPQ lots,6个月accelerated + 12个月real-time只有1批PPQ + 2批earlier process
Container closure systemDP的primary packaging(vial/syringe)锁定后才开始stability还在试不同vial supplier
Storage conditions主储存条件 + ICH accelerated(25°C/60% RH for 2-8°C product;40°C/75% RH for room temp)只有accelerated没有real-time
In-use stability临用稀释后的stability(如diluted in saline)at clinical dose没做in-use
PhotostabilityICH Q1B按整体product评估漏做
Thermal cyclingfreeze-thaw stability for any frozen product漏做
Container closure integrity灭菌后的integrity verification没在commercial container上做

补救路径:stability requires elapsed time——新做需要6-12个月。如果时间紧迫,sponsor可以提供:"stability projection + ongoing stability commitment + accelerated data only",但这种approach让buyer后续承担更多风险,会反映在deal terms(更高的milestone hurdle、更低的upfront)。

红旗7:analytical method validation不完整

MNC commercial site需要transfer methods过去——如果validation report不充分,transfer失败的风险高。期待看到的validation parameters per ICH Q2(R2):

Method必须validation的parameters
Identityspecificity
Assay (concentration)accuracy, precision, specificity, linearity, range
Impurities (quantitative)accuracy, precision, specificity, linearity, range, LOD/LOQ
DAR determinationaccuracy, precision, specificity, linearity(across DAR range)
Free payloadaccuracy, precision, specificity, linearity, range, LOD/LOQ
Aggregate (SEC-HPLC)precision, specificity, linearity, range, LOD/LOQ
Charge variants (cIEF or CEX-HPLC)precision, specificity
Glycan profile (CE-LIF or HILIC-HPLC)precision, specificity, linearity
Cell-based potencyaccuracy, precision, specificity (relative to RS)
Bioassay (binding)accuracy, precision, specificity

中国厂常见缺口:cell-based potency assay的validation报告不充分(intermediate precision不够、analyst-to-analyst variability未characterize);DAR method的validation没覆盖整个DAR分布范围;charge variants的method在不同lots间稳定性未demonstrated。

补救路径:method validation gap可以补——CRO能在3-6个月做完。但如果sponsor的analytical lab本身的data integrity程度低(电子审计trail不全、SOP version不严),buyer会对所有数据可信度生疑。

红旗8:reference standard的qualification和continuity

每个biological product都需要reference standard用于potency assay和identity tests。MNC会问:

  • 当前的reference standard是哪批?什么时候建立的?
  • qualification report是否完整(identity、purity、potency value assignment)?
  • 多大量?什么时候耗尽?next standard的establishment plan?
  • bridging study(current standard与next standard的对比)什么时候做?

中国厂常见缺口:reference standard用的是early clinical batch、qualification report含糊、bridging study没规划。补救路径:planned reference standard transition应当在每5-7年做一次,bridging study要保留充足量overlap。

红旗9:data integrity——FDA/EMA标准

中国厂的data integrity在NMPA inspection中可能足够,但FDA/EMA的要求更严:

FDA/EMA expectation红旗
电子数据系统21 CFR Part 11合规:审计trail、unique user logins、time stamps、disable功能受限共享密码、删除records的能力开放给operator
实验记录ALCOA+原则:attributable、legible、contemporaneous、original、accurate + complete、consistent、enduring、available实验记录不签名、改动没有strikethrough(ALCOA违反)
原始数据保留raw chromatograms、raw spectra、raw assay results保留只保留processed reports
二次核对for critical data,second person review with signature个别analysts独立完成all steps
Data backupoffsite backup + restoration test没有restoration test
Spreadsheet validationcalculation spreadsheets validated per GAMP 5Excel未validate

中国厂的红旗示例:同一台HPLC上,operator能"重做"injection然后只保留"较好"的result——这是classical data integrity violation。FDA inspection曾就此对中国某ADC厂发过Form 483 + Warning Letter。补救路径:在license-out完成前做一次internal data integrity audit,让buyer看到的是已经remediated状态。

红旗10:supplier qualification

ADC生产链涉及mAb intermediate site、drug-linker site、conjugation site、DP fill-finish site,加上critical raw materials(cytotoxic payload、linker chemicals)的suppliers。

供应商类型qualification expectation红旗
mAb intermediate manufacturerfull GMP qualification、quality agreement、change control linkage中国厂内部mAb生产但没有正式quality agreement
Drug-linker manufacturer (small molecule API厂)API GMP per ICH Q7drug linker厂只有Chinese GMP没有FDA registration
Cytotoxic payload supplierGMP + safety data sheet + impurity profileonly one supplier with no backup
Conjugation site (CDMO)GMP + change control + PAI ready中国CDMO没做过FDA inspection
Container closure components supplierextractables/leachables data + GMPonly Chinese suppliers without FDA DMF

补救路径:single-source critical supplier是buyer特别敏感的。Sponsor可以:"identify backup supplier + start qualification但不必完成",明确表态commitment。这比假装"已经dual sourced"被发现要好。

红旗11:regulatory commitments未兑现

Sponsor之前的IND submissions、Type B meeting agreements、NMPA approval letters中是否有"will provide additional data within X months"或"commit to characterize Y by phase 3"的commitments?这些必须能在dataroom中清楚追踪。

DD team会要:

  • 每一次FDA/NMPA/EMA交互的submission letter
  • 每一次agency response(IND clearance、advice letter、IND additional info request)
  • sponsor对每条agency request的response timeline + 实际execution status
  • 如果有commitment延期,timeline justification

红旗:sponsor说"我们已经全部按期完成",但buyer查dataroom发现某个IND amendment中提到的commitment没有对应的delivered evidence。

补救路径:建立一个"regulatory commitment register",按FDA/NMPA/EMA分别列每条commitment的source、deadline、deliverable、status。如有延期,写明reason and revised timeline——透明远比掩盖好。

红旗12:cell line来源和license

中国biotech用了哪个CHO clone?是internal develop的,还是license来的(如Selexis SUREtechnology、Lonza GS、ATUM Leap-In、Sartorius-Albumedix)?

DD问:

  • cell line的license agreement是transferable吗?
  • royalty arrangement在license-out后如何?
  • 如果是internal develop,是否有第三方IP风险(专利infringement)?

补救路径:在license-out谈判前,先做cell line freedom-to-operate analysis,identify所有potential IP risks,并把cell line license的terms透明披露给buyer。如果cell line license不能transfer,buyer可能要求sponsor负担re-establishment cost。

红旗13:site multi-functionality和segregation

中国ADC site如果同时做高活性细胞毒payload和non-cytotoxic biologics,会触发cross-contamination concern:

  • payload manufacturing的segregation(dedicated equipment vs cleanable)
  • HVAC的pressure differential、HEPA filtration
  • product flow在facility中的分离
  • gowning和personnel access controls
  • cleaning validation的residue limits(基于payload的acceptable daily exposure)

EMA对ADC的cross-contamination控制比FDA更严格(EU GMP Chapter 3 + Annex 1对sterile product部分)。补救路径:facility设计若不达EMA标准,sponsor应对"Europe market仅use dedicated suite"或"future build-out plan"做明确承诺。

红旗14:缺Type B/C会议历史和agency feedback

强大的中国创新药CMC dataroom应该包括:

  • pre-IND meeting的minutes(如果做过)
  • Type B EOP meetings(如果到过Phase 2末或Phase 3前)
  • Type C CMC会议
  • 与EMA scientific advice交互(如果有)
  • 与NMPA CDE的pre-IND/IND交互

这些agency feedback对buyer非常valuable——他们能从FDA reviewer/EMA assessor的措辞看出"FDA对这个产品有什么concern"。中国厂没做这些会议("先去做IND再说")的常见后果:buyer对sponsor的regulatory acumen打折,valuation下降。

补救路径:如果之前没做这些meeting,至少做一次pre-NDA meeting或EU scientific advice,让buyer看到sponsor能直接与监管交互。这种会议可以在LOI到signing之间的time window内做完。

红旗15:tech transfer到MNC site的可行性

MNC的commercial site可能位于美国、欧洲或委托给Lonza/Samsung Biologics/Catalent等大型CDMO。tech transfer的可行性(technical transferability)是DD focus:

评估点红旗
Process的scalability当前scale → 2000L或更大当前还是pilot scale,commercial scale未demonstrated
Process的robustnessDOE / process characterization没做characterization
关键设备的transferabilitybioreactor type、purification system、conjugation reactor用了proprietary equipment buyer site没有
Method的transferability用instruments是buyer site都有的吗?用了不通用的analytical platform
Documentation的qualitySOPs、batch records、master batch records中文SOP没有英文版

补救路径:在license-out谈判前期就准备一份"tech transfer feasibility assessment"——sponsor自己列出哪些方面transferable、哪些需要调整、estimated timeline和cost。这种proactive approach比等buyer发现要好。

红旗16-20(速览)

剩下5个常见红旗简述:

红旗16:master batch record缺中英文双版本 — 中国厂的batch records只有中文,buyer的tech transfer team读不了。补救:在license-out前6个月开始bilingualize关键batch records。

红旗17:缺PPQ的pre-defined acceptance criteria — process performance qualification没有事先定的acceptance criteria,PPQ"成功"的判定不可信。补救:PPQ前重新写protocol并retroactively validate(如果数据allow)。

红旗18:EMA-style annual product review缺失 — 中国厂可能有FDA-style APQR但没有EMA-style PQR;或反之。补救:建立适用于FDA + EMA双重格式的APR/PQR。

红旗19:cleaning validation的carry-over和ADE限度计算不严 — 高活性payload的cleaning validation限度如果用一般biologic的approach(10ppm / 1/1000 dose)远远不够。需要用ADE/PDE based limits(按PIC/S PI 053)。

红旗20:environmental monitoring trends不存在或不严 — Annex 1要求的Grade A/B/C/D EM monitoring,许多中国厂只有rolling monthly summary而没有Grade A的continuous monitoring data + alarm system + investigation history。

一份"buyer DD response"的工作流程

收到MNC的CMC questionnaire(200-400问)后,sponsor的应对流程:

阶段时间任务责任人
Day 1-3分类问题把每个问题分类:dataroom里有 / 部分有 / 没有RA Lead
Day 4-7dataroom navigation guide为"dataroom里有"的问题写dataroom location referenceDocument Specialist
Day 8-14起草缺失内容为"部分有"的问题做supplemental memofunction leaders
Day 15-21内部reviewsenior management审阅responsesVP Reg + CMC + QA
Day 22-28提交responses + tracker提交到buyer的data roomRA Lead
Day 29-45回应follow-upbuyer的follow-up questions(一般20-50个)function leaders
Day 46-60site visit准备为可能的site visit准备presentation + tour scriptRA Lead + Operations

关键工具:response tracker(spreadsheet:每条question按ID、category、owner、deadline、response、buyer follow-up状态)。这个tracker后期被buyer的DD team review时也是evidence of process maturity。

一份"data room readiness self-assessment"示例memo

下面这段语言是说明性的,可作为sponsor内部上报CEO的data room readiness assessment参考:

TO: [CEO]
FROM: VP Regulatory & CMC
RE: Data Room Readiness Assessment for [ADC/Bispecific Asset Name] Out-Licensing

We are 6 weeks from the planned data room opening for the proposed
out-licensing transaction with [MNC]. Below is our self-assessment of
CMC data room readiness based on a mock DD performed by [external CMC
DD consultant].

GREEN (ready):
- Cell line history and MCB/WCB documentation
- Drug substance specifications and analytical method validation
- Process A description and validation
- ICH stability program for current commercial process

AMBER (gaps but addressable in 6 weeks):
- Linker stability in human plasma (study ongoing, completion 2026-06-30)
- DAR distribution batch trend (10 batches assembled, additional 6 needed)
- Cleaning validation ADE-based limits (calculation refresh in progress)
- Master batch records English translation (60% complete)

RED (timing risk):
- Process A → Process B comparability (3 PPQ lots not yet released; expect
  release 2026-07-15)
- Reference standard bridging study (will not complete before data room open;
  buyer will see commitment with timeline)
- FDA Type B EOP meeting (scheduled 2026-08-10; advice will be available
  before signing but not at data room open)

Recommended actions:
1. Open data room 2026-07-01 with AMBER items disclosed as "in progress
   with expected completion date"
2. Set buyer expectation that comparability data will be supplemented in
   August
3. Use Type B meeting outcome as a positive signal post-signing to support
   milestone payment trigger

Risk: 5-10% valuation impact from RED gaps, mitigated by transparent
disclosure and strong commitment plan.

这种语言对deal team的价值在于:CEO能直接做risk-aware decision(早开dataroom接受5-10%discount,还是延后6个月等所有都GREEN再开)。隐藏gap让buyer发现的cost通常远大于早期disclose。

与LOI/term sheet的衔接

DD红旗会反向影响deal terms:

红旗严重程度典型buyer响应
1-2个minor gaps(如method validation部分)not material;deal terms不变
3-5个moderate gaps(如comparability未完成)adjustment in milestone schedule(earlier milestones harder)
1-2个major gaps(如data integrity issue、regulatory commitment未兑现)upfront payment降5-15% + 增加quality-related diligence reps & warranties
3+ major gaps或critical issue(如cell line IP风险)walk away或彻底renegotiate

中国sponsor的最优策略:在LOI签署前主动disclose所有known gaps + 提供remediation plan。这种approach让buyer的negotiation leverage降低(因为已经price-in了gap),sponsor保留更多deal value。

参考资源

本文为出海合规研究文章,所引用的ICH、FDA、EMA、NMPA规范内容截至发文日(2026-05-06),最新版本以各官方网站为准。涉及具体license-out交易的CMC风险评估和合同条款,应结合各资产的具体技术profile、临床阶段、deal terms由sponsor的regulatory、CMC、legal、BD团队联合决议;本文不构成法律或商业意见。

AI 助手

你好!我看到你正在阅读「ADC/双抗license-out CMC data room红旗:海外药企尽调最容易抓出的20个质量问题」。有任何关于这篇文章的问题,都可以问我!

由 Gemini 驱动 · 回答仅供参考