2023 White Paper on Recent Issues in Bioanalysis: EU IVDR 2017/746 Implementation/Impact, IVD/CDx/CLIA Approved Assays, High Dimensional Cytometry, Multiplexing Technologies, LBA Tissue Analysis, Vaccine Study Endpoints, Cell-Based Assays for Biomarkers, Cell Therapy and Vaccines (PART 2 – Recommendations on Development & Validation of Biomarkers, IVD, CDx, Cell-Based, Flow Cytometry, Ligand-Binding and Enzyme Assays; Advanced Critical Reagents Strategies)

Abstract

The 17^th Workshop on Recent Issues in Bioanalysis (17^th WRIB) took place in Orlando, FL, USA on 19–23 June 2023. Over 1000 professionals representing pharma/biotech companies, CROs, and multiple regulatory agencies convened to actively discuss the most current topics of interest in bioanalysis. The 17^th WRIB included 3 Main Workshops and 7 Specialized Workshops that together spanned 1 week to allow an exhaustive and thorough coverage of all major issues in bioanalysis of biomarkers, immunogenicity, gene therapy, cell therapy and vaccines.

Moreover, in-depth workshops on “EU IVDR 2017/746 Implementation and impact for the Global Biomarker Community: How to Comply with these NEW Regulations” and on “US FDA/OSIS Remote Regulatory Assessments (RRAs)” were the special features of the 17^th edition.

As in previous years, WRIB continued to gather a wide diversity of international, industry opinion leaders and regulatory authority experts working on both small and large molecules as well as gene, cell therapies and vaccines to facilitate sharing and discussions focused on improving quality, increasing regulatory compliance, and achieving scientific excellence on bioanalytical issues.

This 2023 White Paper encompasses recommendations emerging from the extensive discussions held during the workshop and is aimed to provide the bioanalytical community with key information and practical solutions on topics and issues addressed, in an effort to enable advances in scientific excellence, improved quality and better regulatory compliance. Due to its length, the 2023 edition of this comprehensive White Paper has been divided into three parts for editorial reasons.

This publication (Part 2) covers the recommendations on Biomarkers, IVD/CDx, LBA and Cell-Based Assays. Part 1A (Mass Spectrometry Assays and Regulated Bioanalysis/BMV), P1B (Regulatory Inputs) and Part 3 (Gene Therapy, Cell therapy, Vaccines and Biotherapeutics Immunogenicity) are published in volume 16 of Bioanalysis, issues 9 and 7 (2024), respectively.

Keywords: :

• bioanalysis
• biomarkers
• cell therapy
• gene therapy
• immunogenicity
• vaccine
• WRIB

Abbreviations and Definitions
AAV:	=	Adeno-associated virus
ABC:	=	Antibody binding capacity
ADA:	=	Anti-drug antibody
AI:	=	Artificial intelligence
ALQ:	=	Above the Limit of Quantification
ASO:	=	Antisense oligonucleotide
BAV:	=	Biomarker assay validation
BLA:	=	Biologics license application
BLI:	=	Biolayer interferometry
BLQ:	=	Below the Limit of Quantification
BMV:	=	Bioanalytical method validation
CAP:	=	College of American Pathologists
CAR T:	=	Chimeric antigen receptor T-cell
CBA:	=	Cell Based Assay
CDR:	=	Complementarity-determining region
CDx:	=	Companion diagnostic
CE:	=	Conformité Européenne
CFCA:	=	Calibration free concentration analysis
CLIA:	=	Clinical Laboratory Improvement Amendments
CLSI:	=	Clinical Laboratory Standards Institute
CMC:	=	Chemistry, manufacturing and controls
CoA:	=	Certificate of Analysis
CRISPR:	=	Clustered regularly interspaced short palindromic repeats
CRO:	=	Contract research organization
CSF:	=	Cerebrospinal fluid
CTR:	=	Clinical Trials Regulation (536/2014)
CyTOF:	=	Time-of-flight mass cytometry
ddPCR:	=	Droplet digital polymerase chain reaction assays
DIG:	=	Digoxigenin
DoL:	=	Degree of labeling
DSP:	=	Digital special profiling
ECP:	=	Erythrocyte microparticle
ELISA:	=	Enzyme-linked immunosorbent assay
EMA:	=	European Medicines Agency
EMP:	=	Endothelial microparticle
eQC:	=	Endogenous quality control
ERF:	=	Equivalent Number of Reference Flurophores
EV:	=	Extracellular vesicle
FFP:	=	Fit-for-purpose
FFPE:	=	Formalin-Fixed Paraffin-Embedded
FMO:	=	Fluorescence minus one
FMx:	=	Fluorescence minus one or three
FNA:	=	Fine needle aspiration
FTE:	=	Full-time equivalents
GalNAc:	=	N-acetylgalactosamine
GCP:	=	Good Clinical Practices
GCLP:	=	Good Clinical Laboratory Practices
GLP:	=	Good Laboratory Practices
GOI:	=	Gate of interest
GI:	=	Gastro-intestinal
HMW:	=	High molecular weight
HRMS:	=	High resolution mass spectrometry
IDE:	=	Investigational device exemption
I/E:	=	Inclusion/exclusion
IFU:	=	Instructions for use
IHC:	=	Immunohistochemistry
IMC:	=	Imaging mass cytometry
IND:	=	Investigational new drug
IQ/OQ:	=	Installation qualification/operation qualification
IRB:	=	Institutional review board
ISR:	=	Incurred sample reproducibility
ISS:	=	incurred sample stability
IU:	=	Intended use
IVD:	=	In vitro diagnostic
IVDD:	=	In Vitro Diagnostic Directive (98/79/EC)
IVDR:	=	In Vitro Diagnostic Regulation (2017/746)
KOL:	=	Key opinion leader
LBA:	=	Ligand binding assay
LCMS:	=	Liquid chromatography mass spectrometry
LDT:	=	Laboratory developed test
LLOQ:	=	Lower limit of quantitation
LMP:	=	Leukocyte microparticles
LOB:	=	Limit of blank
LOD:	=	Limit of detection
mAb:	=	Monoclonal antibody
MDB:	=	Multi-domain biotherapeutic
MdFI:	=	Median fluorescence intensity
MESF:	=	Molecules of equivalent soluble fluorochromes
mIF:	=	Multiplexed immunofluorescence
MMO:	=	Mass minus one
MOA:	=	Mechanism of action
MRD:	=	Minimum required dilution
MP:	=	Microparticle
MRE:	=	Minimal number of required events
MRM:	=	Multiple reaction monitoring
MRPS:	=	Microfluidic resistive pulse sensing
MS:	=	Mass spectrometry
MSD:	=	MesoScale Discovery
NAb:	=	Neutralizing antibody
NB:	=	Notified Body
NDA:	=	New drug application
NGS:	=	Next-Generation Sequencing
NME:	=	New Molecular Entity
NTA:	=	Nitriloacetic acid
pAb:	=	Polyclonal antibody
PAD:	=	Peripheral Artery Disease
PBMC:	=	Peripheral blood mononuclear cells
PC:	=	Positive Control, used in an immunogenicity assay
PCR:	=	Polymerase chain reaction assays
PD:	=	Pharmacodynamics
PET:	=	Positron emission tomography
PK:	=	Pharmacokinetics
PMP:	=	Platelet microparticles
POC:	=	Point of care
PTM:	=	Post-translational modifications
PQ:	=	Performance qualification
QC:	=	Quality control
qPCR:	=	Quantitative polymerase chain reaction assays
RMT:	=	Receptor-mediated transcytosis
RNA:	=	Ribonucleic acid
RO:	=	Receptor occupancy
RUO:	=	Research use only
SD:	=	Standard deviation
SEC-MALS:	=	Size Exclusion Chromatography-Multi Angle Light Scattering
SNR:	=	Signal to noise ratio
SOP:	=	Standard operating procedure
SPR:	=	Surface Plasmon Resonance
t-SNE:	=	t-distributed stochastic neighbor embedding
tAb:	=	Therapeutic antibody
TE:	=	Target engagement
Target engagement:	=	Interaction of ligands with their target biomolecules.
Tfh:	=	T-Follicular Helper
TMDD:	=	Target-mediated drug disposition
UK NEQAS:	=	United Kingdom National External Quality Assessment Service
VALID:	=	Verifying Accurate Leading-edge IVCT Development
WRIB:	=	Workshop on Recent Issues in Bioanalysis

Index Part 2

Introduction

SECTION 1 – Biomarkers, IVD and CDx Assays Development & Validation

Hot Topics & Consolidated Questions Collected from the Global Bioanalytical Community

Discussions, Consensus and Conclusions

Recommendations

SECTION 2 – Cell-Based Assays

Hot Topics & Consolidated Questions Collected from the Global Bioanalytical Community

Discussions, Consensus and Conclusions

Recommendations

SECTION 3 – Ligand-Binding/Enzyme Assays and Critical Reagents

Hot Topics & Consolidated Questions Collected from the Global Bioanalytical Community

Discussions, Consensus and Conclusions

Recommendations

References 

1. Introduction

The 17^th Workshop on Recent Issues in Bioanalysis (17^th WRIB) took place in Orlando, FL, USA on 19–23 June 2023. Over 1000 professionals representing pharma/biotech companies, CROs, and multiple regulatory agencies convened to actively discuss the most current topics of interest in bioanalysis. The 17^th WRIB included 3 Main Workshops and 7 Specialized Workshops that together spanned 1 week to allow an exhaustive and thorough coverage of all major issues in bioanalysis of biomarkers, immunogenicity, gene therapy, cell therapy, and vaccines.

Moreover, in-depth workshops on “EU IVDR 2017/746 Implementation and impact for the Global Biomarker Community: How to Comply with this NEW Regulation” and on “US FDA/OSIS Remote Regulatory Assessments (RRAs)” were the special features of the 17^th edition.

As in previous years, WRIB continued to gather a wide diversity of international industry opinion leaders and regulatory authority experts working on both small and large molecules as well as gene, cell therapies, and vaccines. facilitate sharing and discussions focused on improving quality, increasing regulatory compliance, and achieving scientific excellence on bioanalytical issues.

The active contributing chairs included:

Dr. Mitra Azadeh (Ultragenyx), Mr. Mike Baratta (Takeda), Dr. Gopa Biswas (US FDA), Dr. Katherine Block (Genentech), Mr. Mark Dysinger (Alexion), Dr. Seongeun Julia Cho (US FDA), Dr. Isabelle Cludts (UK MHRA), Ms. Kelly Coble (Boehringer Ingelheim), Dr. Vilma Decman (GSK), Dr. Steven Eck (AstraZeneca), Dr. Anna Edmison (Health Canada), Dr. Fabio Garofolo (BRI Frontage),

Dr. Swati Gupta (AbbVie), Dr. Shawna Hengel (Seattle Genetics), Ms. Sarah Hersey (BMS), Dr. Allena Ji (Chiesi), Dr. Wenying Jian (Janssen), Dr. Surinder Kaur (Genentech), Dr. Uma Kavita (Spark Therapeutics), Dr. Christopher Kochansky (Exelixis) Dr. Yi-Dong Lin (Takeda), Dr. Meena (Stoke), Dr. Johanna Mora (BMS), Dr. Rachel Palmer (Sanofi), Dr. Susan Richards (Sanofi) Dr. John Smeraglia (AstraZeneca), Dr. Ivo Sonderegger (Takeda), Dr. Yuan Song (Genentech), Dr. Hiroshi Sugimoto (Takeda), Dr. Matthew Szapacs (AbbVie), Dr. Martin Ullmann (Fresenius Kabi), Dr. Meenu Wadhwa (UK MHRA), Dr. Jian Wang (Crinetics), Dr. Russell Weiner (Takeda), Dr. Long Yuan (Biogen), Dr. Yiyue (Cynthia) Zhang (US FDA).

The participation of major and influential regulatory agencies continued to grow at the 17^th WRIB during its traditional Interactive Regulators' sessions including presentations and panel discussions on:

• Regulated Bioanalysis/BMV and Biomarkers/CDx/BAV: Dr. Mohsen Rajabi Abhari (US FDA CDER), Mr. Abbas Bandukwala (US FDA CDER), Dr. Kimberly Benson (US FDA), Dr. Gopa Biswas (US FDA CDER), Dr. Seongeun (Julia) Cho (US FDA CDER), Dr. Xiulian Du (US FDA CDER), Dr. Anna Edmison (Health Canada), Ms. Dulcyane Fernandes (Brazil ANVISA), Dr. Brian Folian (US FDA CDER), Dr. Akiko Ishii-Watabe (Japan MHLW), Dr. Dany Ivanova (Health Canada), Dr. Sean Kassim (US FDA CDER), Dr. Olga Kholmanskikh (Belgium FAMHP), Dr. Elham Kossary (WHO), Ms. Sonja Kwadijk-de Gijsel, (Netherlands IGJ / EU EMA), Dr. Yang Lu (US FDA CDER), Ms. Tamara Pinkney (US FDA CDRH), Dr. Kara Scheibner (US FDA CDER), Dr. Yoichi Tanaka (Japan MHLW), Dr. Mary Thanh Hai (US FDA CDER), Dr. Yow-Ming Wang (US FDA CDER), Ms. Emma Whale, (UK MHRA), Dr. Joshua Xu (US FDA NCTR), Dr. Li Yang (US FDA), Dr. Yiyue (Cynthia) Zhang (US FDA CDER)

• Biotherapeutic Immunogenicity, Gene Therapy, Cell Therapy and Vaccines: Dr. Mohsen Rajabi Abhari, (US FDA CDER), Dr. Nirjal Bhattarai (US FDA CBER), Dr. Alessandra Buoninfante (EU EMA), Dr. Isabelle Cludts (UK MHRA), Dr. Heba Degheidy (US FDA CBER), Dr. Sandra Diebold (UK MHRA), Dr. Akiko Ishii-Watabe (Japan MHLW), Dr. Mohanraj Manangeeswaran (US FDA CDER), Mr. Christian Mayer (Austria AGES/EU EMA), Dr. Kimberly Maxfield (US FDA CDER), Dr. João Pedras-Vasconcelos (US FDA CDER), Dr. Kara Scheibner (US FDA CDER), Dr. Sophie Shubow (US FDA CDER), Dr. Yoichi Tanaka (Japan MHLW), Dr. Seth Thacker (US FDA CDER), Dr. Omar Tounekti (Health Canada), Dr. Daniela Verthelyi (US FDA CDER), Dr. Meenu Wadhwa (UK MHRA), Ms. Leslie Wagner (US FDA CBER), Dr. Joshua Xu (US FDA NCTR).

The 17^th WRIB included the traditional evening roundtables, which were attended by both industry key opinion leaders (KOL) and regulatory representatives. The extensive and fruitful discussions from these roundtables together with the lectures and open panel discussions amongst the presenters, regulators and attendees culminated in consensus and recommendations on items presented in this White Paper.

A total of 55 recent issues (‘hot’ topics) were addressed and distilled into a series of relevant recommendations. Presented in the current White Paper is the background on each issue, exchanges, discussions, consensus and resulting recommendations.

Due to its length, the 2023 edition of this comprehensive White Paper has been divided into three parts for editorial reasons. This publication covers Part 2 recommendations.

1.1. Part 1A – Volume 16 Issue 9 Month May 2024

1.1.1. Mass Spectrometry, Chromatography & Sample Preparation (6 Topics)

1.	Deuterated Drugs: Implication for Bioanalysis
2.	Quantitation of Nanoparticles and Novel Lipids for mRNA Vaccines and Therapeutics
3.	Clinical Tissue, Tumor and FFPE Biopsy Quantification
4.	Hybrid Assays (IA-MS) - New Applications and Enhanced Complementarity with Conventional Technologies
5.	Mass Spec Targeted Proteomics for Large Protein Panels: Regulatory & Technology
6.	Advanced Mass Spec and Multiplexed Approaches for Biotherapeutics

1.1.2. Mass Spectrometry Assays, Latest Developments, Challenges, & Solutions (6 Topics)

1.	High Efficiency and Quality Non-Regulated Bioanalysis
2.	Quantitative Bioanalytical Methods for Small Molecule Covalent Inhibitors
3.	Quantitative Mass Spectrometry for Chiral Bioanalysis
4.	Biomarkers Novel Approaches for Method Development/Validation
5.	Oligonucleotides Novel Approaches for Method Development/Validation
6.	ADC Novel Approaches for Method Development/Validation

1.1.3. BMV & Regulated Bioanalysis Latest Developments, Challenges, & Solutions (6 Topics)

1.	US FDA/OSIS Remote Regulatory Assessments (RRAs)
2.	Sample Collection, Reconciliation, Chain of Custody
3.	Patient Centric Sampling for Clinical Analysis
4.	Special Issues in the Implementation of ICH M10
5.	Look into the Future of BMV ICH M10
6.	PK/PD Data Exchangeability in and out of China

1.2. Part 1B – Volume 16 Issue 9 May 2024

1.2.1. Input from Regulatory Agencies on Regulated Bioanalysis/BMV & Biomarkers/IVD/CDx/BAV

1.2.1.1. US FDA

• Office of Study Integrity and Surveillance (OSIS): A Day in the Life of Our Remote Regulatory Assessments

• Data Integrity Issues from Recent FDA BIMO Inspections and Remote Regulatory Assessments

• Examination on Overall Acceptable Incurred Sample Reanalysis (ISR)

• Common Deficiencies Related to Method Validation in ANDA Submissions

• Use of Biomarkers in Drug Development: A Regulatory Perspective

• CDRH CLIA Categorization Processes

• Opportunities for Collaborative Innovations Informed by Review Experience in Office of Clinical Pharmacology

• 21st Century Cures: Biomarker Qualification and Analytical Guidance

• PD Biomarker Bioanalysis - Guidance Recommendations and Review Experience

1.2.1.2. Health Canada

• ICH M10 Implementation at Health Canada: Comparative Bioavailability Studies for Chemical Drugs

• Case Studies from Health Canada Review: Comparative Bioavailability Studies for Chemical Drugs

1.2.1.3. UK MHRA

• MHRA Inspection Update

• Implementation of Data Integrity Guidance in the UK

1.2.1.4. Belgium FAMHP / EU EMA

• Integrated CDx and pharmaceutical development in EU regulatory landscape

1.2.1.5. Dutch Youth & Health Inspectorate (IGJ)

• Introduction on the Health and Youth Care Inspectorate

1.2.1.6. Brazil ANVISA

• ANVISA updates on ICH M10 Implementation

1.2.1.7. WHO

• Sharing some of the Significant Deficiencies - CRO Inspections

1.3. Input from Regulatory Agencies on Immunogenicity, Gene Therapy, Cell Therapy & Vaccines

1.3.1. Immunogenicity

1.3.1.1. US FDA

• Tools to Assess Immunogenicity Risk and New Computational Methods

• Updates of OCP Efforts on Evaluating Clinical Impact of Immunogenicity

• Neutralizing Anti-drug Antibody Assays for Biologics- a review of format choices in recent 351 (a) BLA approvals

• Biosimilar User Fee Act (BsUFA) III Regulatory Science

• An Update on Immunogenicity Review Committee Activities

• Recent Advances on Cellular and Gene Therapies

• Is your assay fit for purpose?

• Recent updates on Flow Cytometry and Cell Therapy

• Benchmarking and Improving Indel Calling from Oncopanel Sequencing Data

1.3.1.2. EU EMA

• Vaccine Assay Validation: Regulatory Perspective

1.3.1.3. Health Canada

• Assessment of Qualified Laboratories for Cell and Gene Therapies: Health Canada's approach

1.3.1.4. UK MHRA

• NIBSC Reference Reagents for Flow Cytometry Standardization

1.3.1.5. Austria AGES

• Fast-Track Development and Approval of Covid-19 Vaccines

1.3.1.6. Japan MHLW

• Points to Consider for Using ADA Screening Assay signal-to-noise ratio (S/N) as an Alternative to Titer in Immunogenicity Assessment

• Bioanalysis of AAV vector by qPCR and digital PCR

1.4. Part 2 – Volume 16 Issue 8 Month April 2024

1.4.1. Biomarkers, IVD, CDx Assays Development & Validation (7 Topics)

1.	EU IVDR 2017/746 Implementation and Impact for the Global Biomarker Community
2.	IVD/CDx, CLIA Approved Assay and Regulatory Requirements
3.	What is our Biomarker Assay Actually Measuring?
4.	Free Target Assays for Drug Candidates Targeting Soluble Targets
5.	Biomarkers Development/Validation for Vaccine Study Endpoints
6.	Point of Care (PoC) Assays Development and FFP BAV in Clinical Trials
7.	Fit for Purpose Biomarker Assay Validation (FFP BAV) Challenges for LBA & Mass Spec

1.4.2. Cell-Based Assays (4 Topics)

1.	CBA Novel Strategies for Method Development/Validation
2.	Quality Data Generation in High Dimensional Cytometry
3.	Biomarkers Novel Approaches for Method Development/Validation
4.	Cell Therapy and Vaccine Novel Approaches for Method Development/ Validation

1.4.3. Ligand-Binding/Enzyme Assays & Critical Reagents (6 Topics)

1.	Emerging & Multiplexing Technologies in Bioanalysis
2.	LBA Tissue Analysis
3.	Advanced Labeled Critical Reagents Strategies and Hybridoma Phage Display
4.	Advancements in Enzyme Assays
5.	Novel Modalities Method Development/ Validation Challenges
6.	Single Well Analysis (Singlicate) for ADA Assays

1.5. Part 3 – Volume 16 Issue 7 Month April 2024

1.5.1. Gene Therapy, Cell Therapy & Vaccines (15 Topics)

1.5.1.1. Technologies

1.	The Rise of dPCR
2.	Is ELISpot still the Gold Standard for Assessing Cellular Immunogenicity?
3.	NanoString Assay Development/Validation
4.	NGS Assay Development/Validation
5.	qPCR Assays Development/Validation

1.5.1.2. Immunogenicity

6.	Issues with International Reference Standards for Vaccine Clinical Assay
7.	Anti-AAV TAb Post-Dose Assessment: Develop an Efficient Analysis Strategy
8.	Further Considerations on LNP Immunogenicity
9.	Vaccine Clinical Assays Life Cycle Management
10.	AAV TAb/NAb Assessment
11.	Transgene Immunogenicity Assessment
12.	Vaccine Immunogenicity Assessment
13.	PK & Biodistribution for Replication Competent Viral Vectors
14.	PK & Biodistribution for Virus-vector Vaccine
15.	PK & Biodistribution for CAR-T Cells

1.5.2. Immunogenicity of Biotherapeutics (9 Topics)

1.	ISR for ADA Assays
2.	Strategies to Distinguish between Clinical Pre-Existing Antibodies (PEA) and Treatment Emergent Antibodies (TEA)
3.	Universal & Generic Assays Formats for Immunogenicity Assessment
4.	Immunogenicity Assay/Reporting Harmonization
5.	Immunogenicity Risk-based Approaches, Prediction and Mitigation: Focus on Novel Monitoring Strategies and in vitro Immunogenicity Assessment
6.	Lessons Learned from Immunogenicity Studies/Filings: Sharing Advanced Experience Related to Immunogenicity Strategies/Approaches
7.	Immunogenicity of Complex Biotherapeutics
8.	Cut Point Assessment
9.	Assay Comparability

2. SECTION 1 – Biomarkers, IVD & CDx Assays Development & Validation

Olga Kholmanskikhⁿ, Yow-Ming Wang^g, Sarah Hersey^b, Meenu Wadhwa^d, Katherine Block^a, Abbas Bandukwala^g, Matthew Szapacs^c, Russell Weiner^e, Khader Awwad^f, Francis Dessy^h, Sandra Diebold^d, Sean Downing^e, Xiulian Du^g, Fabio Garofolo^j, Shannon Harris^k, Victor Hou^l, Jennifer Jones^b, Sumit Kar^m, Arvind Kinhikar^o, Ming Li^p, Joel Mathews^q, John Meissen^r, Giane Oliveira Sumner^s, Luying Pan^e, Gerard Sanderink^t, Ingrid Scully^u, Johannes Stanta^v, Yoichi Tanaka^w, Stephanie Vauleon^x, Leslie Wagner^g, Kai Wang^y, Liang Zhu^l

Authors are presented in alphabetical order of their last name, with the exception of the first 8 authors who were session chairs, working dinner facilitators and/or major contributors.

Author affiliations can be found at the beginning of the article.

2.1. HOT TOPICS & CONSOLIDATED QUESTIONS COLLECTED FROM THE GLOBAL BIOANALYTICAL COMMUNITY

The topics detailed below were considered as the most relevant “hot topics” based on feedback collected from the 16^th WRIB attendees. They were reviewed and consolidated by globally recognized opinion leaders before being submitted for discussion during the 17^th WRIB. The background on each issue, discussions, consensus and conclusions are in the next section and a summary of the key recommendations is provided in the final section of this manuscript.

2.2. EU IVDR 2017/746 Implementation & Impact for the Global Biomarker Community: How to Comply with this Regulation

When is a Performance Study with an IVD needed and what is the process required? What are recommendations regarding the criteria for EU health institutions? What are considerations for use of pre-existing local test results in clinical trials? What categories of assays are regulated by IVDR? Are assays validated according to ICH M10, but not CE-marked, sufficient to support clinical trials? For global studies in EU, could modified CE-marked IVDs or ‘in-house’ IVDs be used to support study with documentation and validation? What are regulators expectations regarding germline mutation testing to determine eligibility for clinical trials? What are perspectives on existing genetic test results vs. testing performed for clinical diagnosis? Is there an expectation to have small vs. large gene panels for this purpose? Under IVDR, is there a process in place akin to the US Pre-Sub process, where one can seek advice from the European Medicines Agency (EMA) and notified bodies (NBs) to obtain feedback on the validation strategy for assays used for I/E purposes or as primary endpoints?

2.3. Development/Validation Strategies for Biomarkers & IVD/CDx, LDTs, & Regulatory Requirements & IVD Kits Biomarker Assay Validation (BAV)

Can a CLIA certified Laboratory Developed Test (LDT) in an accredited clinical laboratory be used for prospective selection/stratification of US patients in an US clinical trial? In the US, if a drug label indicates “as determined using an FDA approved test”, does an LDT developed in a CLIA laboratory meet the requirements of FDA approved test? What are key validation parameters for assays used for prospective stratification/selection? What are the requirements when the purpose for an assay is to prospectively stratify or select patients? For new/novel biomarkers in rare diseases, considering that there may be no reference or ‘gold standard’ for analyte measurement, what is the expected level of assay validation required for assays used to determine study subject eligibility (e.g., patient selection)? What is an appropriate strategy/ approach for determining an adequate number of samples to be tested for analytical performance of an assay considering that achieving statistical power may not be possible for rare diseases where a substantial number of samples may not be available? Do the FDA expectations regarding requirements for diagnostic kits mentioned in 2018 BMV guidance also apply to IVD CE-marked kits?

2.4. Biomarker Assays Difference: What is our Biomarker Assay Actually Measuring?

Can results from different biomarker assay/platforms simply be compared to one another? Can they be replaced for one another?

2.5. Free Target Assays for Drug Candidates Targeting Soluble Targets

Should free or total target concentrations be measured? Or both? What's the value of free soluble/shed target assays? Are free assays always needed to monitor target engagement (TE)?

2.6. Biomarkers Development/Validation for Vaccine Study Endpoints: Focus on Reference Standards, LBA Method Development Challenges, BAV & Sample Collection

Do we need to include a reference standard in vaccine clinical assays, and are international reference standards required? How do we cope with interferences in biomarker immunoassays development or validation used for vaccine study endpoints? How does it apply to the characterization of assay linearity, dilutability and selection of the most appropriate diluent? Considering that each vaccine immunogenicity sample is unique in terms of analyte characteristics (polyclonal response against multiple vaccine/pathogen epitopes, several isotypes), how many samples should be studied for the determination of assay linearity limits? How to manage contradictory results between samples? What are the approaches to assess dilutional linearity in vaccine immunoassay setting of assay ranges for biofunctional vaccine immunoassays? What is the impact of sample quality on biofunctional vaccine immunoassay validation? What is the guidance on the ISR for BM assays, especially serology assays for measuring antibody response to vaccines? How do we determine limits of sensitivity (ULOQ and LLOQ) for the immunogenicity assays measuring antibody response to vaccine? What is the consensus on the methodology? Do we want to re-assay ALQ/BLQ samples once a new ULOQ/LLOQ is established? Can we use the first test results if they fall within the new range? What are the challenges of online subject recruitment and data verification, and of using in-home self-capillary blood draw collection for antibody testing?

2.7. Point of Care (PoC) Assays Development & FFP BAV in Clinical Trials

Can patient-centric microsampling devices substitute standard collection devices (e.g., vacutainer)? What additional things should be considered when using a POC test?

2.8. Fit for Purpose Biomarker Assay Validation (FFP BAV) Challenges for LBA & Mass Spec: Exosomes, Low Abundant Protein, Peptides/Small Molecule Biomarkers

What are recommendations for surrogate analyte versus surrogate matrix choice for small molecule biomarkers? What are strategies for normalizing quantitation of peptides in saliva and other non-plasma matrices? What are the regulators' thoughts and suggestions on proteoforms in terms of biological validation and sample analysis? What are recommendations for reference choice, normalization strategy, and quality control for low-abundant EV biomarkers?

3. DISCUSSIONS, CONSENSUS & CONCLUSIONS

3.1. EU IVDR 2017/746 Implementation & Impact for the Global Biomarker Community: How to Comply with this Regulation

The In Vitro Diagnostic Regulation 2017/746 (IVDR) applies since 26 May 2022 in the EU [35]. This regulation replaced In Vitro Diagnostic Medical Devices Directive (98/79/EC) (IVDD) and endeavors to enhance the quality and safety of in vitro diagnostics (IVDs) and medical devices in the internal and global market and seeks to safeguard the health of patients and users by bolstering regulatory measures. This includes increased oversight of notified bodies (NBs), risk classification, and strengthening conformity assessment procedures, performance evaluation, performance studies vigilance, and market surveillance. Previous White Papers in Bioanalysis have discussed IVD from the UK MHRA perspective [26,27] but were limited to UK perspective prior to Brexit. There were extensive case studies presented to examine the IVDR impact and discuss its implementation from the sponsor, bioanalytical laboratory perspective, and to elicit regulatory feedback. Additional regulatory perspective on companion diagnostics (CDx) is provided in the 2023 White Paper in Bioanalysis Part 1.

The main scope of the IVDR is “in vitro diagnostic medical devices,” which are used to assess human specimens for the purposes of treatment management. These IVD devices may provide information about physiological or pathological processes, congenital impairments, predisposition to medical conditions, safety for recipients, treatment response, and defining & monitoring of therapeutic measures. This scope of the regulation includes things like reagents, calibrators, instruments, and software used for these purposes. Specimen receptacles are also considered in vitro diagnostic medical devices. The regulation aims to ensure these devices are safe and effective.

The IVDR introduces new terminology, including a new category for ‘companion diagnostic,’ which is defined in Article 2(7). Per the IVDR, companion diagnostics (CDx) are used for determining whether patients are suitable for a particular treatment with a medicinal product. CDx may be used to identify patients who are most likely to benefit from the corresponding medicinal product or patients likely to be at increased risk of serious adverse reactions as a result of treatment with the corresponding medicinal product, or may help identify patients for whom the therapeutic product has been adequately studied, and found safe and effective. CDx quantitatively or qualitatively measure/determine specific biomarkers to identify these unique patient populations. Biomarkers assessed by a CDx can be present in both healthy subjects and/or patients.

To strengthen the governance and IVDR implementation, the Medical Devices Coordination Group (MDCG) composed of Member States experts and chaired by the European Commission issues procedural guidance and the working group on IVDs provides assistance to MDCG on all IVD-related issues.

Key to the IVDR is the addition of new risk-based classification rules to assign IVDs to one of four risk classes, as shown in Table 1 below, and transition away from mainly self-certification by manufacturers to a mandatory certification by NBs for most classes of devices. Guidance on Classification Rules for in vitro Diagnostic Medical Devices (MDCG 2020–16 rev.2) gives examples for illustrative purposes on how to classify an IVD according to the 7 rules described in the IVDR Article 47 and Annex VIII prior to placing it on the market, making it available on the market or putting into service in the Union. This guidance also depicts a flowchart to help determine whether an IVD is a CDx [36].

Table 1. New risk-based classification of IVDs under EU IVDR 2017/746.

Class	Description	Examples	Certification
A	Low Individual Health Risk, Low Public Health Risk	Sample collection receptacles, Buffers	Self-declaration of conformity (with the exception of sterile devices certified by notified bodies)
B	Moderate Individual Health Risk, Low Public Health Risk	Cholesterol self-testing, Pregnancy self-testing, Hematology, Inflammation, Endocrine	Notified body
C	High Individual Health Risk, and/or Moderate Public Health Risk	Blood glucose self-testing, Cancer testing, Genetic testing, CDx, HLA typing	Notified body
D	High Individual Health Risk, High Public Health Risk	High-risk transmissible diseases (HIV, hepatitis, Covid-19), ABO blood typing	Notified body

As regards the current implementation timelines of the IVDR, lengths of the transition periods from previous Directives to Regulation vary according to the IVD risk class taking into account the intended purpose of the devices and their inherent risks [37]. Note that depending on transition periods, IVDD-compliant, IVDR-compliant, and IVDs may co-exist on the market as under specified conditions. The conformity assessments performed under IVDD remain valid during transition periods and all new devices must comply with the IVDR starting May 2022.

The first case study discussed the current challenges with CE certification. As a result of the IVDR implementation, an estimated 80% of IVDs will now need NB oversight and conformity assessment for CE-marking vs only 20% under the IVDD. While the transition periods for recertification of previous CE-marked IVDs has recently been extended, for conformity assessment and certification by NBs of new IVDs a process taking up to 18–24 months is currently estimated. Assessment of conformity with relevant general safety and performance requirements, and the benefit-risk ratio of the device shall be based on scientific validity, analytical and clinical performance data providing sufficient clinical evidence, including where applicable, relevant data (e.g. from performance studies). Unless it is duly justified to rely on other sources of clinical performance data, clinical performance studies shall be carried out. At this time, such a performance study protocol needs to be submitted both to the Ethics Committee and to the relevant National Authority. The application procedures for a clinical performance study under IVDR and for a clinical trial under Clinical Trials Regulation 536/2014 (CTR) can create additional complexity, as the regulatory landscape in the EU is still evolving and not yet harmonized for IVDs.

To comply with IVDR requirements, essentially, if a patient is an EU member state resident and the assay falls under the IVD definition of the IVDR and has an assigned medical purpose, the assay must either be CE-marked for the intended purpose, or be investigated in a performance study to establish or confirm analytical/clinical performance or be an assay manufactured and used in-house by and EU health institution (covered by the exemption applicable to health institutions in the EU according to the IVDR Article 5.5).

Medical devices produced and used within EU health institutions (in-house devices) to meet specific needs of certain patient groups are exempt from most IVDR provisions, including CE certification by NBs. This exemption applies if the health institution complies with conditions in Article 5(5), such as testing laboratory compliance with ISO15189 or national provisions, as the term “in-house IVDs” is usually used instead of “LDTs” within EU regulatory framework. Detailed guidance on this exemption is provided in MDCG 2023-1 guidance. Health institutions, including hospitals, laboratories, and public health institutes supporting the healthcare system, are covered. For IVDs in medicinal product development and Clinical Trial applications, refer to MDCG 2022-10 guidance on the interface between the CTR and the IVDR.

Sponsor-conducted evaluation of IVDR compliance for their assays used in clinical trials was discussed in case studies. Sponsors are then responsible for IVDR compliance, may act as IVD manufacturers and thus are working together with internal teams CROs, and EU clinical laboratories on ensuring IVDR implementation. Important aspects include but are not limited to whether a clinical trial for an investigational medicinal product (IMP) and simultaneously serving as performance study for IVD, is started prior to IVDR application (i.e, 26 May 2022) or after, whether IVD is used for patient management decisions, regulatory status of the IVD (e.g. CE-marked, in-house) and the types and location of laboratories performing the assay (e.g., CRO located in EU as user of CE-marked IVDs, or CRO is CLIA lab in US). Modifying an existing CE-marked assay, in case of a non-EU health institution use, renders sponsor a legal IVD manufacturer with all relevant provisions. Manufacturer or sponsor of a clinical trial assign the intended purpose for the assay and such intended purpose can define the assay as an investigational IVD and its classification with associated requirements. Examples of compliance status and respective requirements are provided in Table 2.

Table 2. IVDR compliance status and requirements.

Type of Assay	Intended Use	Compliance Status and Requirements
CE-marked IVD	Used according to its intended purpose	Compliant with the IVDR. No action needed.
In-house IVD	Validated at EU health-institution for its intended purpose	Must comply with the IVDR Article 5(5) requirements, such as the relevant general safety and performance requirements set out in Annex I and apply the relevant MDCG guidance.
Modified CE-marked IVD (not CE-marked anymore)	Modifications to the CE-marked IVD used within or outside of the initial intended purpose	Must comply with the IVDR (e.g. Annex I). Provide the justification of modifications, performance study to be conducted, where applicable as IVD manufacturer or use as an ‘in-house IVD’. Validation with added calibrators and QC to meet BMV guidance may be needed.
Lab Developed Test (US)	Validated to meet performance requirements in US for its intended purpose, CLIA certified laboratory	Must comply with the IVDR for global study with EU sites or using a global CRO. The exemption provisions of Article 5(5) are only applicable to EU health institutions, and thus not to US labs. Performance study may be needed.

It was shown to be important to collect and use data starting from early clinical stage study for the purpose of further IVD/CDx performance evaluation. The need for developing a CDx is assessed based on the data generated from early clinical development studies. Data from the clinical trials conducted as performance study are used for submission for conformity assessment of CDx by a NB. For a new therapeutic modality, the IVD/CDx is oftentimes not available in early clinical development stage, and it may be necessary to select CRO to develop and validate assays according to regulatory guidance on bioanalytical method development and validation. Overall, it is important to evaluate CRO and assay compliance status.

Finally, global development perspectives from the bioanalytical laboratory on experience with use of IVDs in US and EU, specifically anti-AAV assays for exclusion of patients in clinical trials were shared. A case study was shown for the journey from an anti-AAV assay from preclinical use to IVD use in a clinical trial. The preclinical assay used the same assay format and reagents as the planned clinical trial assay. In the early clinical trial, the anti-AAV assay was needed for patient selection and post dose characterization (assessment of seroprevalence and seroconversion). Thus, assay validation was performed to enable IDE submission in US and CE marking in UK/CA with assay transfer to a specimen testing laboratory. For conducting a pivotal trial with medicinal product finalized version of the candidate CDx assay is recommended. Challenges and risks during this process can include unexpected design changes, rendering CE/UKCA certification no longer valid, PMA Manufacturing and Analytical Module Deficiencies, and temporal misalignment of the notified body certification procedure with marketing authorization application procedure resulting in delay to commercial launch of the medicinal product.

As lessons learned from this case study, in early clinical development, measurements for exploratory endpoints or when no patient management decisions are taken based on assay results can leverage LDT (in US) or in-house IVDs (in EU). For assays requiring a higher regulatory standard (e.g., patient management decisions), an assay adequately validated in a clinical laboratory and meeting the specific regional regulations requirements for use in clinical trials is needed.

As noted above, the definitions of the IVDs and their classification differ between EU and US. For the purpose of comparison, types of assays and their examples under FDA classifications are presented in the Table 3 below.

Table 3. Types of assays and their examples under FDA classifications.

Type of Assay	Purpose of the Assay	Example
Companion diagnostic (CDx)	Essential for safe and effective use of the corresponding therapeutic product	Herceptin^® with HercepTest^™, Zolgensma^® with MEBCDX AAV9 test (Japan only) Roctavian with AAV5 Detect CDx^™
Complementary diagnostic	Identifies a subset of patients that may respond better to the drug. Not essential for the safe and effective use of the drug	Imfinzi^® with Ventana PD-L1 (SP263) assay
Class Companion Diagnostic or Class Complementary Diagnostic	Can be used for multiple products within a therapeutic class	Cobas^® EGFR Mutation Test v2 with Tarceva^®, Tagrisso^®, Iressa^®, Gilotrif^®, and Vizimpro^® for the treatment of EGFR mutated NSCLC
In vitro diagnostic (IVD), 510(K) CLEARED	Assay used to diagnose, monitor, or prevent disease. This category includes both CDx and assays not linked to a therapeutic product	Total immunoglobulin E (IgE) for allergy testing, AdvanSure^™ SARS-CoV-2 IgG (RBD) for SARS-CoV-2 immune response testing
Laboratory developed test (LDT)	Not commercially available as a distributed kit. Can only be performed at laboratory where developed. Can be used for patient management decisions and clinical end points	AAV9 antibody tests at Athena Diagnostics (US), at ImmunoSpot (US), and Viroclinics (EU)
Bioanalytical assay	Exploratory clinical trial endpoint. Not used for patient management decision	PK, PD, ADA, Biomarker assay, evaluated retrospectively.

This table was originally published in ‘Preexisting antibody assays for gene therapy: Considerations on patient selection cutoffs and companion diagnostic requirements’ by Braun, Lange, Schatz et al. (2024). Available at: https://www.sciencedirect.com/science/article/pii/S2329050124000330.

This table was modified and reused with permission from the publisher, licensed under the Creative Commons Attribution – NonCommercial-NoDerivs (CC BY-NC-ND 4.0). https://creativecommons.org/licenses/by-nc-nd/4.0/

Based on these case studies and user questions, recommendations for several aspects of the IVDR were proposed by regulators, sponsors, and bioanalytical laboratories. First, there was review and clarification of the categories of assays that are regulated by the IVDR: CE-marked IVDs, IVDs that undergo performance evaluation (not yet CE-marked for the intended purpose that is being evaluated or CE-marked for a different intended purpose), ‘in-house IVDs’ manufactured and used in EU health institutions and not CE-marked. The scope of application of this Regulation should be clearly delimited from other legislation concerning products, such as medical devices, general laboratory products and products for research use only (Recital 7 of the IVDR). According to the IVDR Article 2(45), “a device intended to be used for research purposes, without any medical objective, shall not be deemed to be a device for performance study”. Thus, research use only (RUO) assays used without any medical objective and an assigned in vitro diagnostic medical purpose are out of the IVDR scope. Compliance with an international regulatory guidance such as ICH M10 is not within the requirements specifically laid down in the IVDR; its principles can be used to prove compliance (e.g. with Annex I of the IVDR) but IVDR provisions are applicable and prevail (e.g., for performance study submission). For global clinical trials with IMPs conducted also in EU, modified CE-marked assays or in-house IVDs can be used provided compliance with the IVDR (per above).

There was also reinforcement of the three main options for a specific case of assays used for biomarker-based selection of patients in EU: use of a CE-marked assay, submission of a clinical performance study (performance study route) if a CE-marked assay is not used for its intended purpose or not CE-marked, or implementation of not CE-marked assay as in-house IVD in an EU health institution. Therefore, it is necessary to understand when a performance study (as defined in Article 2(42) of the IVDR and regulated by Article 57) is needed and what is the application or notification process required, also in case of additional requirements applicable for certain performance studies according to Articles 58(1), 58(2), 70(1) and 70(2).

Per the IVDR, ‘interventional clinical performance study’ means a clinical performance study where the test results may influence patient management decisions and/or may be used to guide treatment. According to Article 58 (1), such a study, as well as a study in which surgically invasive sample-taking is done only for the purpose of the performance study or where the conduct of the study involves additional invasive procedures or other risks for the subjects of the studies, shall, in addition to meeting the requirements set out in Article 57 and Annex XIII, be designed, authorized, conducted, recorded and reported in accordance with this Article and Articles 59 to 77 and Annex XIV of the IVDR. Article 58(2) refers to performance studies involving a CDx (e.g. not yet CE-marked or CE-marked for a different or be investigated in a performance study to establish or confirm intended purpose), while Article 70 is dedicated to performance studies regarding devices bearing the CE-marking either within (involving additional burdensome and/or invasive procedures) or outside the scope of its intended purpose.

The submission documentation package for a performance study application is referred to in the Article 66(1) and includes the clinical performance study plan (CPSP), Investigator Brochure (IB), and other supporting documentation, and it must be submitted to each member state with sites where EU patients will be enrolled. Additional national Member States requirements may apply for the application documentation content. The MDCG 2022-19 guidance is listing and providing templates for performance study application/notification documents under the IVDR and the MDCG 2022-2 guidance outlines general principles of clinical evidence and provides guidance on the continuous process of performance evaluation for IVDs.

Recommendations were given regarding the criteria for EU health institutions as per MDCG 2023-1 guidance. There was agreement that better clarification is needed regarding what constitutes an EU health institution in addition to the definition in this guidance. An appropriate quality management system (QMS) and compliance with ISO 15189 are required and additional applicable national provisions may be required.

The use of pre-existing local testing results was proposed as an option for enrollment in clinical trials. It was recommended that local testing should comply with rules for in-house device testing while central testing is preferred in many instances. In addition, confirmation with central testing may be required with submission of a performance study for an IVD, dependent on each individual use case.

Similarly, perspectives on use of existing genetic test results vs. testing performed for clinical diagnosis including germline mutation testing for eligibility in clinical trials were discussed. The recommendation was to perform testing in compliance with the IVDR (e.g. Article 4) regarding confirmation of pre-existing test results. There is a move toward whole genomic sequencing (WGS), but there are implications for patients due to incidental findings, variants of uncertain significance. Patients tested with a genetic test must be provided with all relevant information on its nature, significance and implications. They must be given appropriate access to counselling in cases where a test provides information on the genetic predisposition for medical conditions and/or diseases that are generally considered to be untreatable. There is no specific provision in the IVDR for use of small vs. large gene panels, respective IVDs should comply with the IVDR.

Finally, there was discussion whether there is a mechanism for obtaining advice from EMA and NB, similar to US Pre-Submission process, to get consensus opinion on validation strategy for assays used for Inclusion/Exclusion or as primary endpoints. Discussions are ongoing but NBs cannot do any consulting according to the legislation and there are currently no structured interactions. Sponsors are recommended to use regular channels for scientific advice.

3.2. Development/Validation Strategies for Biomarkers & IVD/CDx, LDTs & Regulatory Requirements & IVD Kits Biomarker Assay Validation (BAV)

The development and validation of biomarker assays along the clinical development life cycle, from translational research, early clinical trials, and through pivotal clinical trials with biomarker assays for subject enrollment, have been a perennial topic of discussion. Understanding the complexity of regulatory frameworks associated with laboratory developed tests (LDTs per the FDA a form of IVD), IVD, and CDx. These topics have been extensively discussed in the 2020–2022 White Papers in Bioanalysis [27,31,32]. If US patient-specific results from an assay(s) are reported, then the facility (i.e. laboratory) performing the IVD testing must hold the appropriate CLIA certificate (42 U.S.C. § 263a(b)). Other recommendations include engaging with regulatory agencies early to plan for the specific purpose case and required validation package for a biomarker, and the importance of bridging studies late in development and an early emphasis on ensuring adequate quantity and storage of clinical samples to enable that bridging.

There was additional focus and case studies on CLIA requirements supporting the prior recommendations. The CLIA regulations stipulate the conditions that all laboratories must meet to be certified to perform IVD testing on human specimens from. Laboratories may obtain one of the following CLIA certificates, depending on the type of testing they perform: Certificate of Waiver (CoW), Certificate for Provider-Performed Microscopy Procedures (PPMP), Certificate of Registration, Certificate of Compliance (CoC), and Certificate of Accreditation (CoA). Under CLIA, test systems are categorized into three complexity levels: waived, moderate complexity, or high complexity, depending on how difficult they are to perform. The more complex the test is to perform, the more stringent the requirements for the laboratory performing the test. Most CLIA requirements are the same for all non-waived testing, including requirements for facility administration, quality systems, proficiency testing (PT), and inspections. There are additional personnel requirements for high complexity testing. LDTs are considered high complexity tests. In accordance with the CLIA regulations at 42 C.F.R. § 493.3(b)(2), research laboratories that test human specimens but do not report patient-specific results for the diagnosis, prevention or treatment of any disease or impairment may be exempt from CLIA certification requirements.

CLIA requirements for high complexity testing must be applied to LDT assay validation. The validation for LDT assays must include accuracy, precision, analytical sensitivity, analytical specificity, reportable range, stability, reference interval, and any other performance characteristic required for test performance in the laboratory that intends to use it. These are similar considerations as ICH M10, and BMV, but it is recommended to use CLSI guidelines as a starting place since other bioanalytical guidelines are focused on PK assays. CLSI guidelines are also a great starting place for submitting a pre-sub to the FDA including validation plans.

There are many other requirements if the assay is used as an investigational device to prospectively stratify and/or select patients, as indicated in 21CFR Part 812 (labeling, monitoring, conducting in accordance with GCPs, etc.). Assays being deployed as “investigational use only devices (IUO)” need to meet the requirements of 21 CFR Part 812 (Investigational Device Exemptions). Under IDE regulation, an IDE needs to be submitted, or an abbreviated IDE is required unless exempted under 21 CFR 812.2I. Investigational devices are assessed to be either a non-significant risk (abbreviated IDE requirements) or significant risk (full IDE requirements) and applies to gene therapy as well. Non-significant risk devices are devices that do not pose a significant risk to human subjects. Non-significant risk devices are still required to meet abbreviated 21 CFR 812 (812.2(b)), including labeling, monitoring, etc. A significant risk device presents a potential for serious risk to the health, safety, or welfare of a subject [38].

In a case study, an immunoassay for an FDA approved IVD biomarker assay was validated within a CLIA certified laboratory for a different purpose. The analytical validation data was submitted to the FDA to support a Ph 3 trial. The FDA feedback received emphasized that it is very important to perform precision studies using clinical specimens from the intended use population (i.e., relevant disease state) rather than using contrived samples consisting of known positive controls spiked into a sample matrix. Accuracy and precision should use a risk-based approach, i.e., more stringent criteria when risk is high, and the precision test should use CLSI EP05-A3 instead of CLSI EP15-A3.

This led to discussion of validation requirements for patient selection assays for novel biomarkers in rare diseases, considering that there may be no reference or ‘gold standard’ for analyte measurement. There was agreement that it is necessary to discuss a proposal with health authorities (e.g., the use of contrived samples vs. true patient samples). Using real world evidence (RWE) datasets is encouraged, if applicable. No One-Size-Fits all approach is possible, and validation depends on the intended purpose, technology, etc. For example, for AAV/gene therapy patient selection, cut points and differentiation of values close to cut point should be a key consideration. Various levels of QCs near the cut point, and repeated precision testing are typically required. It is also important to plan and discuss with agencies the adequate number of samples to be tested for analytical performance of an assay considering that achieving statistical power may not be possible for rare diseases where a substantial number of samples may not be available.

In addition, the process of CDx development was reviewed. Advice was shared based on experiences supporting prior recommendations. Companion diagnostics are unique in vitro diagnostic tests used to ensure the safe and effective use of a therapeutic product. Timelines for both the therapy and the device must be considered. Reliability of results is integral to minimizing risk to the patient & study. There are multiple paths to successful CDx development and approval, and thus it is important to gather inputs from across the business and create a strategy that meets the needs for studies with line of sight to commercialization. Crucially, starting and planning early is key. Sponsors are encouraged to develop a diagnostic strategy early including discussion with regulators to align on requirements (Pre-submission meetings, scientific advice meetings etc). Finally, experience has shown the importance of selecting a technology/assay partner that meets the needs of the program.

Finally, there was further discussion of assays developed for a primary or secondary endpoint in a clinical trial. The 2021 White Paper in Bioanalysis recommended that the validation package required for FDA cleared IVDs and LDTs for endpoints depends on the purpose of use. A case study was discussed where a biomarker in human cerebrospinal fluid was to be used as a secondary endpoint in Ph 2 and Ph 3 trials assessing changes from baseline. The IVD kit planned to be used was validated for serum. Therefore, a full assay validation according to 2018 BMV was required.

This study led to a discussion of whether FDA expectations regarding requirements for diagnostic kits mentioned in 2018 BMV guidance also apply to FDA cleared IVD/CE-marked kits. There was agreement that the 2018 BMV guidance that covered the use of diagnostic kits requires replacement with a more focused and clearer biomarker-specific guidance. Often the requested evaluations set forth in the 2018 BMV guidance are not possible to achieve in practice. Manufacturers will not release proprietary information that is often needed to understand how the assay was developed, how materials and processes are controlled, the level of validation, and key performance characteristics. Therefore, it is incumbent on the sponsor to identify what experiments, if any, need to be done to supplement what was done by the manufacturer for the specific purpose. Important scientific questions include whether samples are in a different matrix, if the disease state evaluated is different, whether the assay range is sufficient to encompass the range of samples anticipated, and the stability conditions.

If the purpose is a primary/secondary endpoint, the recommendation is to consider discussing with the regulatory agency the level of validation rigor needed to support this usage. Coming prepared with scientific justification is necessary, understanding what experiments can be done and which are nearly impossible. While diagnostic assays are highly characterized over many years before they are approved/cleared, sponsors cannot always solely rely on the data contained in the kit insert especially if the test is being deployed for an alternate indication or use case. It is important that performance of the assay is established within the laboratory where testing will be conducted. If the purpose for the assay is the same, and instructions for use are being followed, a separate validation may not be needed. However, it will be important to verify the assay performs as intended within the laboratory conducting the testing.

Updated recommendations and feedback on sponsor questions for LDTs, IVDs, and CDx development were provided. First, it was asked whether a Laboratory Developed Test (LDT) or “CLIA Approved Assay” developed in a CLIA certified laboratory can be used for prospective selection/stratification for US trials. Regulatory Authorities stated that it is unclear what is meant by a “CLIA Approved Assay”, and further explained that an assay may be cleared or approved by the FDA through the premarket notification process (e.g. 510k, PMA) for in vitro diagnostic (IVD) use. Following FDA clearance/approval, IVD test systems, which frequently include an assay, are then CLIA categorized based on the level of complexity (i.e. how difficult they are to perform) as described above. An approved assay more adequately describes an assay that has undergone FDA premarket review. Therefore, it was recommended that the statement “CLIA Approved Assay” not be used since this terminology is incorrect.

The CLIA regulations at 42 CFR 493.1253(b)(2) require laboratories that modify an FDA-cleared or approved test system, or introduce a test system not subject to FDA clearance or approval (including methods developed in-house (i.e., LDTs)), or use a test system in which performance specifications are not provided by the manufacturer, to establish the performance specifications for the performance characteristics (listed above) for the test system, as applicable, before reporting patient test results.

The next discussion question was whether an LDT in a CLIA certified laboratory meets the FDA requirements of an FDA approved test on a US drug label “as determined using an FDA approved test”. If the US drug label indicates, “as detected by an FDA-approved test”, the only tests acceptable are tests formally cleared/approved by the FDA, which cross-reference the therapeutic. Use of other assays for making prescribing decisions would be considered in the strictest sense, “off-label usage”.

3.3. Biomarker Assays Difference: What is our Biomarker Assay Actually Measuring?

Biomarkers and biomarker data can help speed up drug development and guide clinical decisions. This enthusiasm has led to a remarkable increase in the number of biomarkers being measured in clinical trials for biomarker discovery and exploratory to make drug development decisions. Consequently, biomarkers and biomarker assay platforms are too diverse to establish a one-size fits all set of regulations. The consistent dogma has been that biomarker development and validation should be FFP. However, as captured in previous White Papers in Bioanalysis [30], there has been confusion around the term “COU” (context of use). Therefore, it was recommended that it is important to have a unified approach for implementing a successful exploratory biomarker strategy during drug development rather than just a checklist of assay attributes to achieve. A focus back to the “what, why, where, who, and how” method development basic questions is critical to understand prior to starting development and validation of any bioanalytical assay to ensure accurate, reproducible, and biologically/ clinically meaningful biomarker data can be produced [39].

In this regard, additional case studies were discussed to demonstrate how to improve reproducibility in discovery and exploratory biomarker assays. The lack of reproducibility in biomarkers has been widely published, highlighting the need for improvement [40–43]. Case studies demonstrated the impact and need of choosing the right matrix for discovery biomarker assay reproducibility. In the case of a biomarker associated with the brain for example, the levels and regulation of proteins may be different depending on where samples are taken (e.g., brain tissue vs. CSF). Accumulation of proteins may be different depending on the sampling site. Preclinical work to understand how concentrations of biomarker proteins differ between the CSF and brain as well as between various parts of the brain are key to neuroscience/neurology drug development. The brain tissue is the site of action but brain tissue is often not attainable in clinical trials and thus CSF or blood must be used as a surrogate.

Other case studies have shown wide variation in biomarker results when comparing data between mass spectrometry and ELISAs, or even between different ELISA kits. This demonstrates the importance of trying to determine the proteoforms of the protein that are being quantitated during assay development and validation. It is recommended to use orthogonal methods, as well as the same reagents, if possible, when switching between platforms. Since a reagent or platform change could change what the assay is measuring, some form of bridging exercise would be needed if moving from assay A to B. Prior recommendations strongly supported that biomarker discovery and validation needs to be integrated and planned up front. This cannot be done as an afterthought. In addition, it is important to take the time to understand what you are measuring and since relying only on the manufacturer's product insert may not be adequate to develop an analytical validation plan. The time needed for a thorough evaluation to improve the quality of biomarker data should also be communicated to internal teams during drug development.

Results from different biomarker assays on the same platform (e.g., two ELISAs) or across platforms (e.g., MSD vs Olink) cannot simply be used interchangeably. A reagent or platform change could change what the assay is measuring and therefore, some form of bridging exercise would be needed if moving from assay A to B. If the assay/platform changes, it is incumbent on the scientist to justify, likely through orthogonal methods, what they are measuring. Reference standards should be the starting point to anchor across assays. True biomarker (endogenous protein) reference standards are the exception.

3.4. Free Target Assays for Drug Candidates Targeting Soluble Targets

Free target, bound target, and total target biomarker assays are complex sets of assays that are used to assess pharmacokinetics, pharmacodynamics, and target engagement (TE). Due to their complexity, they have been extensively discussed in multiple White Papers in Bioanalysis [30,32]. It is not common practice to offer all three types of assays (free, total, and complexes) for a TE study. Previous studies have focused on the challenges of analyte overestimation in these assays. As such, free and complex target measurement needs knowledge of binding kinetics to assess assay feasibility and reagent development. Free target overestimation and analytical variability tolerance should be defined for study decisions. Total assays are less susceptible to the issue of complex dissociation as this can be tested during assay validation. When developing assays, rapid separation of bound and free target is preferred due to complex dissociation concerns. Free target overestimation and analytical variability tolerance should be defined for supporting study decision making.

The 2022 White Paper in Bioanalysis also provided in-depth case studies of PK/PD modeling for these assays. The generation and use of internal modeling and simulation tools for free and total assay development was recommended to inform assay experimental feasibility and build PK/PD models that allow the use of total target and free drug to predict free target (suppression/target coverage). Pre-requisites for these models include in-solution Kd values and highly accurate binding partner concentrations.

There continues to be a debate on whether to measure soluble/ shed “free” (unbound to the drug) and/or “total” target (both unbound and bound to the drug) concentrations. Free target levels may better reflect optimal dosing, but free assays often yield BLQ results post-dosing. Total target assays with modeling can estimate free target as an alternative. The target assay format may also depend on the PK assay format. For example, a total target assay may be appropriate when used in combination with a free PK assay but may be less desirable if developing a total drug PK assay due to high levels of soluble target. The distinct challenges for free and total soluble target assays, as well as recommendations for their characterization were reviewed. New case studies for free and total target assays were also discussed to highlight their potential value in aiding in the understanding of PK data.

Additional case studies were shown with a subcutaneous Phase 1 dose escalation study using free and total assays [44]. Lessons learned were that a clear understanding of the target biology and its interaction with the therapeutic is an important factor in the design of the bioanalytical methods intended to measure either total or free target concentrations. Generation and careful characterization of assay reagents is critical to ensure the development of accurate/robust methods to measure target concentrations (either total or free). For free target assays, special care should be taken in the selection of capture reagents to minimize disruption of the sample equilibrium, with dissociation of target: drug complexes present in the sample and, consequently, over estimation of free target levels. For total assays, drug interference (as well as interference from endogenous binding partners) may need to be overcome to ensure accurate quantitation of total target levels.

For monitoring target engagement prior recommendations were supported. There was agreement that total target should be measured. Due to the many challenges in developing free target assays, it is necessary to consider if it is needed and why. If there is a biological reason or clinical results that can't be explained by the total data, a free assay may be beneficial. It was also recommended to consider if a modelling approach can be used (use total data to predict free). There are cases where PK/PD modeling was accurate, but this may not be the case for all targets. Overall, there should be discussions between clinical pharmacologists advocating for the utility of the data and the bioanalytical team evaluating the feasibility/degree of difficulty in the assay development to ensure a reliable method can be developed, and meaningful data generated.

3.5. Biomarkers Development/Validation for Vaccine Study Endpoints: Focus on Reference Standards, LBA Method Development Challenges, BAV & Sample Collection

Amidst the COVID-19 pandemic, the vaccine community embarked on an extensive discussion of BAV for clinical vaccine assays used for endpoints [26]. For serology assays, there was emphasis on the need to demonstrate consistency over the life cycle of the assay and a phased development approach (setup, qualification, validation). Detailed recommendations were given for linearity, specificity, LOD, LLOQ, precision, and stability. Trending analysis, proficiency panels, and critical reagent characterization were also described. More recommendations were recently provided such as how to define ULOQ with clinical samples as they become available, parallelism testing, and reference standard requirements [32]. Finally, harmonization efforts for vaccine LBA validation were described which are still ongoing.

While this harmonization is ongoing, case studies were shared on method development challenges for vaccine studies. The first study detailed how to cope with interference and apparent lack of linearity in validation of immunogenicity assays. Contrary to many bioanalytical assays for therapeutics (e.g., PK assays) for which linearity can be demonstrated by spiking increasing amounts of the analyte (e.g., the drug) in a negative matrix, for biofunctional vaccine assays generally considered as complex, demonstration of linearity is challenging. Linearity is traditionally demonstrated by spiking an increasing amount of a positive sample in a negative matrix (sample). In this case, both analyte concentration and matrix concentration vary. The ability to dilute a sample is demonstrated by diluting a positive sample in an assay buffer. If the same amount of analyte is not measured at the same level in matrix and in assay diluent, it is concluded that matrix is interfering with measurement. However, it does not automatically mean that the assay is not performing well. Indeed, it is critical to make the distinction between components interfering with the assay itself (e.g., biotin present in a sample can interfere in a streptavidin-based assay) and components interacting together to elicit the targeted response measured in the assay (outcome of composite biological reactions). Specifically, for biofunctional assays, which measure the level of activity (e.g., neutralizing, bactericidal) and not directly the concentration of an analyte, the measured activity can be the result of several interacting components that can have a positive, negative, or a combinatorial effect. A consideration of these interactions is important because they also occur in vivo.

These studies on linearity and interference led to the recommendation that assay linearity is considered as a valid surrogate of assay accuracy for vaccine LBAs and biofunctional assays. Linearity measures activity elicited by the sample and not concentration. As mentioned above, selection of an appropriate matrix for use as diluent is critical. For binding assays, it is recommended to use negative matrix (since it contains a mixture of substance representative of real samples). For biofunctional assays, however, problems can arise from factors (e.g., cross-reactants, heterophilic antibodies etc.) present in negative matrix which can potentiate/interfere with the activity of samples. Because the relative proportion of contributing factors from the negative matrix (sample) and the positive sample vary across the dilution series, it can result in a non-linear response. Since in this scenario, expected measurement is not predictable, use of antibody-depleted serum (often commercially purchased so a cautious approach is urged) as diluent is recommended as it is more likely to generate a linear dilution series (relative proportion of contributing factors kept constant across the series).

The number of samples to determine linearity limits varies among manufacturers. Besides sample availability and other practical considerations (Phase 1 vs Phase 2) that may apply in specific cases, each sample is unique in terms of analyte characteristics (heterogeneous polyclonal response against various vaccine/pathogen epitopes, isotypes, different affinities/avidities etc.). It is therefore recommended to justify the rationale for sample numbers tested and the approach taken. Samples over the range of the assay can be fundamentally different samples which reflect the evolution of the immune response (e.g., different affinities/avidities/amounts), therefore samples covering the assay range are needed. Hence, using one single high responder sample diluted down to the LLOQ is not recommended. If contradictory results between samples are shown, there was consensus to analyze data and investigate the cause further to explain the discrepancies. It is plausible that acceptance criteria may still be met. If validation fails, investigation of the root cause and impact and how it can be rectified is useful. Importantly, perfect samples (true negative) do not generally exist and the selection of the most appropriate matrix to demonstrate linearity should be part of the study.

Another case study discussed the benefits of using human reference standards for vaccine assays. Anti-serum-based reference standards for evaluating antibody responses are used to minimize the variability across different versions of assays, ensure uniformity in the assigned activity, and allow comparisons between laboratories. In the study, a reference standard composed of human sera from vaccinated individuals was implemented to facilitate maintenance of assay performance and comparability of results across multiple years and different target populations (e.g., infants and the elderly).

This led to a discussion on if there is a need to include a reference standard in vaccine clinical assays and whether international reference standards (IS) are required. This was a contentious topic with varied opinions expressed by different stakeholders. For reference standards, the context is important and how they are used in relevant internal comparisons (e.g., assay platforms/studies vs immune correlate assays). They can minimize assay variability and provide consistent data. They are also useful for binding assays but biofunctional assays may not always benefit from reference standards (as precision is impacted). A majority view among industry was that while control of assays is important, reference standards and IS are not always required. Serology standards for IS can help towards standardizing assays in the long term, however since sera from naturally infected individuals are often used for IS, they are not always representative of sera from vaccine-treated individuals and can provide results which are not meaningful. This is certainly true given the diversity in available vaccines (e.g., vaccine types, mode of action, antigens) for some pathogens. Consequently, IS may not be applicable in all cases. Generally, IS are globally helpful in harmonizing data where feasible, and can be helpful for regulatory bodies to understand relative behaviour across different vaccines. However, it is important not to overinterpret results. e.g. two assays developed to support two vaccines with different mode of action, could aim measuring two different antibody subsets from the IS, making the comparison irrelevant.

Updated recommendations were developed for LBA method development and validation challenges for vaccine study endpoints. Guidance was given on questions for ISR, ULOQ, and LLOQ for BM assays, especially serology assays for measuring antibody response to vaccines. There was an agreement that at present, there is no guidance for ISR traditionally performed for PK assays. The need for ISR for vaccine assays is questionable. Instead, there is a focus on the use of QCs and proficiency panels (low, mid, high QCs) for stability studies and trend monitoring. The LLOQ with acceptable accuracy and precision needs to be defined (qualification) and confirmed later during validation with samples near and below this level. Similarly, ULOQ may need to be redefined using clinical samples and prior recommendations should be used [32].

Another related topic was the reassessment of above or below the limit of quantitation (ALQ/BLQ) samples once a new ULOQ/LLOQ is established. This is dependent on accuracy/linearity and the confidence in what the assay is measuring (accuracy and precision). In cases where the ULOQ is not high enough, re-assessment is required with clinical samples depending upon clinical endpoints. If nothing else has changed, there is no need to re-assay samples; however, if LLOQ has changed, then re-evaluation may be needed.

The final discussion was about the practical challenges of online subject recruitment, data verification, and of using in-home self-capillary blood draw collection for antibody testing. There was agreement that recruitment was easier during pandemic, but it is now more challenging to enroll patients, and for patients to self-test. Shipment and transport to and from patient's home can add significant logistical challenges. Training, verification of training, and compliance with micro-sampling is another challenge. Pilot testing prior to roll-out should be considered.

3.6. Point of Care (PoC) Assays Development & FFP BAV in Clinical Trials

Biomarkers can be measured using samples collected at a patient's home or bedside and sent to a lab for analysis. However, results may be delayed for this approach compared to using a point of care device that can test a fresh sample on a cartridge providing the results more quickly. Point of care testing and microsampling have been extensively discussed for applications in mass spectrometry, LBA, and biomarkers [32,34]. Consistently, it has been recommended that operational and pre-analytical considerations are important and a fit for purpose assay development and validation criteria are needed.

Notably, ICH M10 BMV does give the flexibility to allow patient centric sampling and for any innovative technologies including new microsampling plan. Since multiple parameters affect biomarker results, both pre-analytical (sample processing, extraction of dried material, loading of sample on cartridges) and analytical (sample volume, normalization, analytical platform selection, assay sensitivity etc.) factors need to be considered and adapted as part of validation. The COU of the study also matters so that the right validation steps can be carried out. This becomes increasingly important when the measurement is meant for human diagnostics, disease progression, treatment efficacy, and safety monitoring.

Additional case studies of microsampling in applications for vaccine trials and biomarker assays were discussed. For biomarkers, routine clinical assays using vacutainers were converted to fingerprick collections with a stepwise validation, collaboration with the central lab, and consideration of dead volumes in automation. For CoV-2 vaccine trials, challenges were shared on the implementation of at home sample collection using capillary blood draw devices for antibody testing and for monitoring duration of anti-SARS-CoV-2 antibodies post vaccination. Innovative methodologies that use decentralized procedures to assess real-world immunogenicity are crucial to providing real-time information on the durability of COVID-19 vaccine responses. These data can help inform public health decisions around vaccination campaigns and assess the emergence of variants of concern (VOCs) along with the need for updated booster vaccines.

These case studies led to discussion of whether patient-centric microsampling devices can substitute for standard collection devices (e.g., vacutainer). With the increasing number of biomarkers tested, prior recommendations were confirmed suggesting that patient-centric microsampling devices (e.g., Tasso) may be able to replace a standard collection device. It should be noted that since the different devices collect samples from different compartments (venous vs capillary), any microsampling collection device cannot simply be used interchangeably with a vacutainer without first conducting experiments that demonstrate the different collection methods are comparable (to avoid past challenges in the early days of dried blood spots). Patient training on how to properly use the device is critical. Like diagnostic tests, PoC tests cannot simply be used without additional experimentation to understand what the test manufacturer has done (e.g., differences in matrix, disease state, stability, etc.). It is likely that the intended use of the PoC test (to distinguish between healthy and disease) is not the same as how it's used in drug development. Hence to avoid a false negative or false positive, the PoC test needs to be shown equivalent to the gold standard assay.

3.7. Fit for Purpose Biomarker Assay Validation (FFP BAV) Challenges for LBA & Mass Spec: Exosomes, Low Abundant Protein, Peptides/Small Molecule Biomarkers

FFP BAV has been perennially discussed in prior White Papers in Bioanalysis and the C-Path White Paper [25,27,32,45] among others. The standing recommendation has been that it is crucial to determine suitable analytical validation criteria based on the endogenous analyte and by measuring the analytical error and biological variation in the assays [25]. The end goal is to determine (relative) accuracy, precision, analytical measurement range, parallelism, specificity, selectivity, and stability of the methods. However, since no formalized regulatory guidance has been published, continued discussion of challenging biomarker assays is necessary to provide recommendations on FFP BAV for these cases.

Both surrogate analyte and surrogate matrix approaches need to be evaluated during method development and pre-validation. In the case where no appropriate surrogate matrix is found, one may consider the surrogate analyte approach. In both cases, the accuracy of the selected approach must be tested using well-characterized samples of the appropriated target population during validation.

In this regard, new case studies were discussed for FFP BAV of exosomes, low abundance proteins, and peptides/small molecule biomarkers. For small molecules, one study utilized LCMS and the recommendations from recent white papers to perform development and FFP validation of a sensitive biomarker assay for K2EDTA plasma of Leukotriene B4 and 5-Oxo-Eicosatetraenoic Acid (5-oxoETE) [46]. FFP challenges that had to be overcome included choice of control matrix since the commercial K2EDTA plasma displayed substantially different chromatographic interference than K2EDTA plasma collected via clinical study protocol. A custom pooled K2EDTA plasma lot was created to match clinical trial collection. Additionally, processed sample stability assessment of the isotopically labeled 5-oxoETE surrogate analyte met target stability criteria @ 2–8°C. However, analysis of the 5-oxoETE biomarker revealed increasing abundance over time in authentic matrix only, suggesting formation of the biomarker or a co-eluting interference @ 2–8°C. This study led to the recommendation that surrogate analyte and matrix choice for small molecules must be evaluated on a case-by-case basis, noting that there are some common solutions (e.g., dilution, affinity purification). A DOE data driven approach comparing multiple methods together during development may be considered [47]. Patient super pools can be an option harmonized with commercial samples similar to the case study above.

Another case study used LCMS and immunoassays to develop and validate biomarker neuropeptide panels in saliva and plasma. It was shown that when assessing a panel of peptides, development of sample extraction methods (immunocapture or solid-phase extraction) needs to be extensively optimized to allow for efficient recovery of the various neuropeptides while minimizing any peptide degradation or non-specific adsorption.

Neurodegenerative protein biomarkers, like RGMa or Tau, consist in multiple proteoforms by undergoing splicing or post-translational modification (e.g., >100 for Tau). These proteoforms may be patient- and disease state-dependent, and the assessment of those proteoforms is important to understand the protein biomarker dynamics. Recommendations were given by the expert panel on assessing proteoforms in terms of biological validation and sample analysis. There was agreement that proteoform specific assays can be considered in addition to using literature for biomarker of interest to build body of evidence over time.

The recommended strategies for normalizing quantitation of peptides based on the above and other case studies in saliva and other non-plasma matrices was normalizing to the total protein. For urine, normalizations to creatinine and osmolarity are options. Multiple methods should be evaluated during development, and validation should further be evaluated during validation to allow the selection of the optimized normalization method. Inter-patient and intra-patient changes are important to assess when evaluating normalization methods. Ethnic diversity, disease biology, and gender may also affect quantitation.

A final discussion was based on a case study which reviewed advancements in biomarker assays for low-abundant proteins in biofluids including extracellular vesicles (EVs). For low-abundant proteins, Simoa continues to be supported as the recommended method for ultrasensitive protein measurement, consistent with prior recommendations EVs [32]. Sample enrichment methods were also compared. Recent advancements were shown for EVs utilizing Simoa [48]. Prior recommendations for low-abundance EVs, including reference choice, normalization strategy, and quality control are discussed. EVs and pre-analytical considerations have been extensively discussed in previous White Papers in Bioanalysis [27,30,32] and continue to be a work in progress. Despite growing experience, consistency of enrichment strategies is not yet clinically viable. Further discussions and updates at future WRIB meetings are needed to support this new field and provide guidance to the industry.

4. RECOMMENDATIONS

Below is a summary of the recommendations made during the 17^th WRIB:

4.1. EU IVDR 2017/746 Implementation & Impact for the Global Biomarker Community: How to Comply with this Regulation

• In Vitro Diagnostic Regulation 2017/746 (IVDR) became applicable in 26 May 2022 in the EU. All new assays now need to comply with the IVDR.

• CE-marked IVDs, in-house IVDs manufactured and used within health institutions, IVDs undergoing performances studies are covered by the IVDR, while research use only (RUO) assays without any medical objectives and an assigned in vitro diagnostic medical purpose are out of its scope.

• Use of assays for biomarker-based selection of patients in EU has currently 3 main options:

  • Use assay CE-marked for the intended purpose;

  • Submit an application for a performance study with submission package including CPSP, IB and other documentation (performance study route) if assay is not CE-marked or CE-marked outside the scope of the intended purpose;

  • Use an in-house IVD in an EU health institution when specific IVDR requirements can be met and apply the relevant MDCG guidance.

• Performance studies regulated by the IVDR and requiring submission of application include:

  • Interventional clinical performance studies, studies with surgically invasive sample-taking only for the purpose of the performance study, studies with additional invasive procedures or other risks for the subject;

  • Performance studies involving CDx (e.g. interventional clinical performance studies investigating IVDs intended to be used as CDx; for performance studies using left-over samples only notification to the competent authority is required);

  • Performance studies with CE-marked devices used within the scope of their intended purposes which involve additional burdensome and/or invasive procedures;

  • Performance studies with CE-marked devices used outside the scope of their intended purposes.

• ICH M10 principles for assay validation (applicable generally to PK assays) can be used when showing compliance of biomarker assays with IVDR, but such regulatory guidance is not within requirements of the IVDR and its provision and guidance should be followed.

• Non-CE-marked assays such as in-house IVDs (or LDTs) can be used to support a study with documentation and validation provided they are part of a Performance Study submission and comply with IVDR requirements.

• An EU health institution must be authorized in their country to perform patient testing. They do not have to treat patients, but the majority of activities must be in pursuing health.

  • QMS and compliance with ISO 15189 or where applicable national provisions are required.

• Use of pre-existing local test results for eligibility in clinical trials is an option.

  • Local testing should comply with rules for in-house device testing

  • Confirmation with central testing can be required with submission of a performance study for an IVD.

4.2. Development/Validation Strategies for Biomarkers & IVD/CDx, LDTs, & Regulatory Requirements & IVD Kits Biomarker Assay Validation (BAV)

• If the clinical trials require examination (i.e., laboratory testing) of samples taken from the human body of US citizens for a medical decision (e.g., prospective selection/stratification), then the facility (i.e. laboratory) performing the LDT/IVD testing must hold the appropriate CLIA certificate.

• Regulatory Authorities stated that it is unclear what is meant by “CLIA Approved Assay”.

  • An assay may be cleared or approved by the FDA through the premarket notification process (e.g., 510k, PMA) for in vitro diagnostic (IVD) use. Following FDA clearance/approval, IVD test systems, which frequently include an assay, are CLIA categorized based on the level of complexity (i.e., how difficult they are to perform) as described above.

  • An “approved assay” more adequately describes an assay that has undergone FDA premarket review. Under CLIA, FDA cleared/approved IVD assays are CLIA categorized.

  • Therefore, it was recommended that the statement “CLIA Approved Assay” is incorrect. CLIA requires analytical validity of LDTs, and assay is validated for use in that laboratory but not reviewed by FDA.

  • Assays being deployed prospectively for stratification or selection as “investigational use only devices (IUO)” need to meet the requirements of 21CFR Part 812 (Investigational Device Exemptions). Under IDE regulation, an IDE submission, or abbreviated IDE requirements are required unless exempted under 21 CFR 812.I.

    • Investigational devices are assessed to be either a non-significant risk (abbreviated IDE requirements) or significant risk. This also applies to gene therapy.

  • If the US drug label, indicates “as detected by an FDA-approved test”, the only tests acceptable are tests formally cleared/approved by the FDA, which cross-reference the therapeutic.

    • Use of other assays for making prescribing decisions would be considered “off-label usage”.

  • As applicable, precision (intra/inter-assay), accuracy, reagent stability, linearity (if applicable), specificity, sensitivity (LOD/LLOQ), etc. should be included for analytical validation of assays used for patient enrollment.

    • For rare diseases with limited reference standards and clinical samples, discuss the proposal with health authorities (e.g., use of contrived samples vs. true patient samples, number of samples required, etc.).

• The use of IVD cleared/CE-marked kits for bioanalytical biomarker assays (e.g., primary or secondary endpoints, but not used for patient selection) often requires FFP validation and additional documentation.

  • 2018 BMV guidance that briefly mentioned the use of diagnostic kits needs further clarification; often the requested evaluations set forth in the 2018 BMV guidance are not always feasible but should be followed whenever possible.

  • If a primary/secondary endpoint assay, consider discussing with the regulatory agency the analytical validation plan or scientific justification why additional validation is not needed.

4.3. Biomarker Assays Difference: What is our Biomarker Assay Actually Measuring?

• For exploratory biomarker strategy, it is important to have a unified approach for assay development and validation. The exploratory biomarker strategy should avoid a simple checklist approach of assay attributes to ensure data is accurate, reproducible, and clinically meaningful.

  • It's important to take the time to understand what is being measured.

  • Relying only on the manufacturer's product insert may not be adequate to gather the analytical data for the biomarker.

  • This may necessitate educating clinical teams on the time needed to do a thorough evaluation of platforms, and reagents.

• It is typically not possible to compare results from different assays on the same platform (two ELISAs), or assays on different platforms (MSD vs. Olink) to another.

  • A reagent, platform, or matrix change could change what the assay is measuring and therefore, some form of bridging exercise would be needed if moving from assay A to B.

  • If the assay/platform changes, it is incumbent upon the scientist to justify, likely through orthogonal methods, what they are measuring.

  • Reference standards (recombinant or endogenous proteins) should be the starting point to anchor across assays. True biomarker (endogenous protein) reference standards are the exception.

4.4. Free Target Assays for Drug Candidates Targeting Soluble Targets

• For target engagement, recommendations from prior White Paper in Bioanalysis were reinforced.

  • Total target should be measured. Due to the many challenges in developing free target assays, one should consider if it is needed and why.

  • If there is biological reason or clinical results that can't be explained by the total data, a free assay should be attempted.

  • Consider if a modelling approach can be used (use total data to predict free).

  • Communication with internal teams is necessary to balance the difficulty of free target assay development with the usefulness of the data for clinical pharmacology.

4.5. Biomarkers Development/Validation for Vaccine Study Endpoints: Focus on Reference Standards, LBA Method Development Challenges, BAV & Sample Collection

• Regarding the need for reference standards for vaccine clinical assays, context is important and exactly how are they used. They can minimize assay variability and provide consistent data.

  • Biofunctional assays may not always benefit from reference standards (as precision can be impacted).

  • International standards can help standardize assays in the long term, however they are not always representative.

• Assay linearity is considered as a valid surrogate of assay accuracy for Vaccine LBAs and measures activity, not concentration.

  • In biofunctional assays, factors present in negative matrix can potentiate/interfere with activity of samples.

  • Biofunctional vaccine assays are complex, and linearity is challenging. Sample selection and assay diluent (antibody depleted serum) are critical.

• ULOQ may need to be redefined using clinical samples and prior recommendations should be used.

• The need for ISR for vaccine assays is questionable. Instead, there is a focus on QCs and proficiency panels (low, mid, high QCs). Stability studies can be conducted and trend monitoring undertaken.

4.6. Point of Care (PoC) Assays Development & FFP BAV in Clinical Trials

• Prior recommendations were confirmed that patient-centric microsampling devices (e.g., Tasso) may be able to replace a standard collection device.

• A microsampling collection device cannot simply be used interchangeably with a vacutainer without first conducting experiments that demonstrate the different collection methods are comparable.

• PoC tests cannot be used in drug development without additional experimentation to understand what the test manufacturer has done (e.g., differences in matrix, disease state, stability, etc.)

4.7. Fit for Purpose Biomarker Assay Validation (FFP BAV) Challenges for LBA & Mass Spec: Exosomes, Low Abundant Protein, Peptides/Small Molecule Biomarkers

• There was continued recommendation that BAV must determine suitable analytical validation criteria based on the endogenous analyte to determine (relative) accuracy, precision, analytical measurement range, parallelism, specificity, selectivity, and stability of the methods.

• For small molecule biomarkers, surrogate analyte and surrogate matrix choice are key.

  • Evaluate on a case-by-case basis but noting there are some common solutions that can be employed (e.g., dilution, affinity purification). Patient super pools are another option.

• To normalize peptides in saliva and other non-plasma matrices, total protein is commonly used.

  • Multiple methods should be evaluated during development. Validation should further evaluate the chosen normalization method. It is important to evaluate inter-patient and intra-patient changes.

• Suggestions on proteoforms (splicing and post-translational modification) in terms of biological validation and sample analysis include using proteoform specific assays or supporting decisions with literature.

• Currently, Quanterix's Simoa is recommended to be one of the leading methods for low-abundant biomarker development. However, with significant unmet needs, Simoa needs to be further developed to realize its full potential, expand analyte availability, and increase instrument accessibility. Proper sample enrichment could also help to enhance the detection of low abundant protein in biofluids.

• It is agreed that EVs hold promises as valuable biomarkers, it is challenging to bring them to the clinic due to their heterogeneity and high technical challenges. Despite lacking consensus, SEC was recommended for EV enrichment, and nano-sized flow cytometry was suggested for EV characterization. Reasonable calibrators were also suggested.

• The ability to enrichment EVs consistency is not yet clinically viable but is a work in progress.

5. SECTION 2 – Cell-Based Assays

Steven Eck^aa, Yi-Dong Lin^e, Mitra Azadeh^z, Vilma Decman^y, Sandra Diebold^d, Xiulian Du^g, Polina Goihberg^ab, Enrique Gomez Alcaide^x, Christele Gonneau^ac, Michael Nathan Hedrick^b, Gregory Hopkins^ad, Sumit Kar^m, Jakob Loschko^u, Megan McCausland^ae, Luis Mendez^af, Sarita Sehra^ag, Erin Stevens^ah, Yongliang Steve Sun^b, Shabnam Tangri^ai, Paul C Trampont^l

Authors are presented in alphabetical order of their last name, with the exception of the first 5 authors who were session chairs, working dinner facilitators and/or major contributors.

Author affiliations can be found at the beginning of the article.

6. HOT TOPICS & CONSOLIDATED QUESTIONS COLLECTED FROM THE GLOBAL BIOANALYTICAL COMMUNITY

The topics detailed below were considered as the most relevant “hot topics” based on feedback collected from the 16^th WRIB attendees. They were reviewed and consolidated by globally recognized opinion leaders before being submitted for discussion during the 17^th WRIB. The background on each issue, discussions, consensus and conclusions are in the next section and a summary of the key recommendations is provided in the final section of this manuscript.

6.1. CBA - Novel Strategies for Method Development/Validation

For clinical use assays, CLSI H62 suggests that elimination of markers from panels to simplify assays for an analysis is permissible without additional validation under specific conditions. Do we think this should be acceptable practice for GxP validated assays? What are the recommendations regarding sample number and distribution in generating assay comparability studies for situations such as adding markers to panels? Is it a reasonable practice to share established reference type ranges between assays if those assays have demonstrated comparability? Are there situations where adding a marker might not require a full validation? Is using existing validation datasets normally adequate to validate changes in data analysis such as alternate gating strategies including possibly performance validation of new measurements within the data? In clinical use assays, samples collected outside the method instructions are often still tested as an irreplaceable sample. Are there instances where such a deviation would make sense in regulated bioanalysis by flow cytometry and what should be done in such situations? What is the real advantage of supervised automated gating and automated data analysis workflows in the clinical arena (outside of retrospective data analysis)? What could be the minimal requirements or acceptance criteria from the regulatory authorities to accept and validate the use of these tools in clinical trials? Is there enough evidence and studies showing robustness and reproducibility on the results generated by automated data analysis workflows?

6.2. Quality Data Generation in High Dimensional Cytometry

How to validate AI/ML analysis? How to report clinical data from AI/ML analysis? What are the standards for this type of analysis that the industry should consider? What are strategies for harmonization of high-dimensional flow cytometry panels to enable comparability between different clinical studies?

6.3. Biomarkers Novel Approaches for Method Development/Validation

Should we implement both indirect/competitive and direct/bound RO formats for a clinical trial (uses different reagents)? How to select a sample matrix for an RO assay, peripheral blood, PBMCs or tissues? What are new approaches for membrane target expression quantification with flow cytometry? What are challenges in exosome applications as biomarkers using flow cytometry? For new platforms such as spectral flow cytometry in clinical trials what is the pathway to successful platform adoption? What are the responsibilities of the instrument manufacturer (manufacturing specifications, IQ/OQ) versus the end user (PQ)? What role do the instrument and reagent manufactures play in proactively developing standardization reagents and approaches (hard-dye beads vs fluorochrome specific particles)? What reference controls to use for spectral flow cytometry in clinical setting? How to choose appropriate validation specimens and target populations for LLOQ with high dimensional flow cytometry?

6.4. Cell Therapy & Vaccine Novel Approaches for Method Development/ Validation

What are strategies for cell-mediated immunity assays (CMI) and flow cytometry based intracellular cytokine staining (ICS) assay development to measure T Cell immune responses? What are workflows for assay qualification and for validation of CMI and ICS assays? What are the analytical considerations for different points in the CAR T manufacturing and treatment process? What characteristics of those samples are important to understand the potential success of the CAR T product and treatment?

7. DISCUSSIONS, CONSENSUS & CONCLUSIONS

7.1. CBA Novel Strategies for Method Development/Validation

The Clinical and Laboratory Standards Institute (CLSI) H62 is the first comprehensive document that builds on the foundational consensus papers published over the last decade for flow cytometry. Since then, White Papers in Bioanalysis have provided supporting recommendations for validation and specific assay development considerations that have arisen in cytometry [27,30,32,49]. Recommendations have been given for validation protocols with pre-specified acceptance criteria, biological variability, addition of unplanned analytes, accuracy assessment, and sensitivity determination. Specific challenges previously identified have included analysis of samples outside established stability, PK/PD analysis supporting Project Optimus, rare matrices, iterative validation, and critical reagent characterization [32,50].

Additional case studies addressing major difficulties and challenges in flow cytometry validation and clinical use were discussed and support prior recommendations. One case study reviewed a possible revision of dendritic cell (DC) gating strategy to better assess DC subsets. Since no changes to staining and cytometer acquisition were proposed, existing datasets (i.e., fcs files from validation) were used in a retrospective ‘electronic validation’ to qualify the newly defined analytes generated through the revised gating strategy. The new gating strategy was refined on non-validation samples, a validation amendment plan incorporating the new analysis was written, and performance of analysis of the new measures was assessed relative to their intended use.

Another case study illustrated the use of dried antibody cocktails as a potential solution for error-prone manual master mixing in clinical flow cytometry. Dried antibody mixes were shown to be stable long term at room temperature. Regarding the continued discussion of clinical sample analysis outside of stability windows, case studies were shown using stabilizing tubes (e.g., Cyto-Chex BCT, TransFix, Smart Tubes, etc.) to increase sample stability, enabling a more flexible workstream due to the presence of mild stabilizers in the whole blood collection tubes. However, it was observed that stabilization buffers must be tested empirically to ensure key biomarkers are not impacted by the fixative nature of chemicals present in these commercially available collection tubes. Other proposed solutions were the use of alternate platforms such as ChipCytometry and epigenetic immunophenotyping. ChipCytometry allows retrospective interrogation but needs evaluation of the effect of PBMC isolation and fixation on cell populations and key biomarkers. Epigenetic immunophenotyping correlated with flow cytometry measurement of cell populations and eliminates the need for a viable cell population, overcoming logistical challenges. However, one limitation is that it does not provide protein expression information at the cellular level.

For instances where sample analysis must occur outside stability (whether this occurrence would be due to the lack of any of the solutions described above or due to a simple logistics issue or delay), prior recommendations were supported as follows: since the sample was already collected and tagged as an imprecise analyte in database, it can be tested with documentation of a deviation, analyzed outside of validated stability, being conscious of how the data reported and analyzed in the context of clinical trials will be used to prevent incorrect interpretation.

The expert panel then further discussed new challenges and potential solutions in flow cytometry assay development and validation. First, it was asked if the removal of markers from existing GxP-validated panels would require additional validation. The consensus was a risk-based approach should be used to determine what the appropriate comparability assessment required to make the assay fit-for-purpose (FFP). -It was noted that the approach would depend on where the marker to be removed fits in the gating strategy used for data analysis (downstream gating with no effect on current gating, for informational use only) and depends on absence of impact on reportable (i.e.: removal of pan-leucocyte CD45 marker may have a more serious impact that removal of a more downstream marker of activation on lymphocyte subsets such as PD-1). It is important to show non-negative impact to panel performance through the comparison of FMO and/or FMX to a fully stained panel to confirm comparability and consider risks to things such as prior stability determinations. Ideally, the comparison of FMO and/or FMX to a fully stained panel would have been assessed during the initial panel development and optimization phase, to allow for changes such as marker elimination without additional testing.

The process for adding markers to an existing validated panel was also discussed. There was a consensual agreement that, in a majority of cases, adding a new marker constitutes a new assay (significant change) requiring a FFP validation. In some cases (e.g., testing clones on same fluorochrome), there may be method development data to support significant change or non-significant change decision. Comparability assessments should be performed if needed to identify changes in values and changes in distributions. Quality controls should be updated to account for the newly added marker(s).

The use of existing datasets to validate changes in data analysis such as alternate gating strategies, including performance validation of new measurements within the data was also discussed. There was consensus that in many instances, this should be acceptable. It was added that non-validation data could be used to adjust analysis and then new in silico analysis should be applied to the validation dataset.

Finally, there was continued discussion of the advantages of supervised automated gating and unsupervised automated data analysis workflows in the clinical arena (outside of retrospective data analysis). Prior recommendations from white papers were confirmed across the board. These automated methods are usually not critical but can be highly helpful: they are unbiased, save time (combinations of markers, off-hour analysis), and reduce the number of analysts and QC steps. Experience so far is that they require substantial investment before becoming useful. Hence, they can be helpful in Phase 3 trials or for widely used assays where extent of sample testing offsets the investment required to train the automation algorithm. The utility of auto gating is improving which will only increase in the future. For validation to be acceptable by regulatory authorities, at minimum, users should demonstrate utility of assay (FFP), perform software validation, and use in analytic validation. In general, for most assays, there is not sufficient robustness and reproducibility data for unsupervised automated data analysis, though some limited examples exist of utility (Minimal Residual Disease assessment for example), and data is accumulating across the board.

7.2. Quality Data Generation in High Dimensional Cytometry

Full spectrum cytometry allows scientists to examine various markers in higher detail at the single-cell level. With this technique, the number of detectors and filter sets does not limit panel design as in conventional flow cytometry, generally permitting higher numbers of stains to be used within assays. Careful selection of available fluorophores is still required to control spread and ensure that adequate separation and sensitivity are achieved. Spectral cytometry, along with advances in reagents (for conventional and spectral cytometry) have generally made higher “plexing” of assays more readily achievable. Datasets generated from such assays have similar size and scope issues as are encountered with other high plex technologies, such as mass cytometry. Previous recommendations have been provided on the early challenges of high dimensional cytometry such as situations to develop the assays, what instrument(s) to leverage, sample handling, marker/sample stability, data handling and storage, data analysis (hierarchical vs. AI/ML), computing environment, and reporting. For example, it was previously recommended to thoroughly evaluate expertise within the lab/CRO for availability of validated panels, equipment, and data reporting capabilities [32]. Discoveries that are not biased by prior knowledge can be made by unsupervised analysis, but software should be validated, or findings should be confirmed using a validated manual approach.

Supportive case studies were discussed for implementing high dimensional flow cytometry in clinical testing. For example, users are standardizing the core immunophenotyping markers for commonly used PBMC cell types and pathways to aid standardization and harmonization of readouts used for clinical sampling. This includes standardization of sample handling, reagents, antibody cocktails, staining protocols, instrument setup and harmonization, gates, and data analysis. Antibody titration is also being standardized [51]. Other case studies were presented for quality data generation with automated gating in global trials.

These harmonization efforts led to discussion of strategies to enable comparability between studies. It was agreed that it is difficult to compare across sponsors and trials without harmonization. CLSI H62 and other peer-reviewed best practice publications have recommendations to facilitate harmonization [52], but it was noted that commercially available materials for harmonization and standardization may not adequately represent behavior of fluorochromes used in different assays.

There was additional discussion on how to validate and report data obtained from AI/ML analysis. There was agreement that typically, clinical questions are already pre-defined. Most often, AI/ML approaches are used for signal finding (e.g., clinical data correlation). This can be followed by manual confirmation through conventional gating. Overall, it is important to retain data traceability. One approach is to define the performance characteristics of the manual and automated gating using the validation dataset to determine whether assay performance is FFP. Examples of this approach can be found in genomics and digital pathology which also use AI/ML analysis. It is important to segregate training/association discovery datasets and testing/confirmation datasets. Validation datasets can serve as a testing dataset in some instances to help understand/demonstrate reproducibility in highly similar samples (such as from precision and stability testing). In other cases, separate datasets may be required to evaluate reproducibility of association finding (e.g., population X and clinical response). Auto gating and AI/ML approaches require significant investment and large sample numbers for validation and the importance of upfront data curation to eliminate poor quality stains should not be underestimated.

7.3. Biomarkers Novel Approaches for Method Development/Validation

The rise of new technology used for spectral cytometry, imaging cytometry, or mass cytometry have permitted the use of these platforms into novel applications to serve translational and clinical sciences for drug development. Some of applications include receptor occupancy (RO) measurement and exosome analysis.

Regarding RO assays in peripheral patient blood, previous recommendations have discussed total, bound, and free receptor measurements [30,32]. There is a consensus that typically two out of the three receptors are measured, and the third is inferred. The total receptor measurement uses a non-competitive detection antibody to the target or takes the approach of unlabeled drug saturation of the test sample followed by detection with an anti-drug detection reagent. The bound receptor measurement uses an anti-drug detection reagent. When anti-drug detection reagents are used to assess bound or total (post saturation of the test sample) receptors, special attention should be given to the titration of the detection reagent to minimize background. And finally, the free receptor measurement uses a competitive antibody that ideally has a similar affinity to the drug or uses directly labeled drug [30,32].

Reagent availability and characteristics were reinforced as the main determinant of the selection of RO assay formats. Prior recommendations were also supported for RO development and validation [53]. One new case study showed measurement of PD-1 RO using PBMCs in a clinical trial [54]. However, prior recommendations were supported that while the matrices may vary, blood is still most commonly used for measurement of peripheral RO.

Case studies were shown using flow cytometry measurement to support gene therapy studies. For example, cell-specific simultaneous detection of transgene mRNA and protein with PrimeFlow^™ - methodology was discussed [55]. Additional applications of imaging cytometry were also shown in measuring mRNA endosomal escape with imaging flow-cytometry. The expert panel discussed whether new approaches are available for membrane target expression quantification in tissue with flow cytometry. There was agreement that in non-liquid tissues it is difficult to control dissociation (need to validate separately). Spike in controls can be used, but similar to other technologies, there is no perfect way with flow cytometry. Caution was encouraged in interpretation due to pre-analytical considerations (e.g., biopsy contamination like hemodilution). More experience is needed for further recommendations.

Another still novel application of flow cytometry that was discussed is the measurement of extracellular vesicles (EVs). EV isolation methods, pre-analytical characterization, organ specific EV quantification, and detection methods by LBA, mass spectrometry, and flow cytometry have been discussed in previous White Papers in Bioanalysis. There was previous agreement that flow cytometry is the most common method to count and characterize microparticles (MPs), which are very small EVs. The main challenge in analyzing MPs is their size, which is below one micron. New instruments with high sensitivity can improve the resolution and detect smaller nanoparticles, which can help measure MPs more accurately.

A case study was shown supporting the use of flow for EVs and MPs by determining the expression level and enzymatic activity of CD73 on EVs from cell lines and plasma. EVs were isolated by size exclusion chromatography and detection of CD9 and CD73 on EVs was performed by on-bead flow cytometry. Despite successful case studies, there was agreement that 1) there is a higher bar for exosomes (due to complex procedures), 2) there is a lack of optimal isolation methods for use in flow cytometry and 3) that further standardization is needed. Size of the particles remains challenging for most flow cytometers. Specific newer flow cytometers (e.g., Attune^™ and CytoFLEX cytometers) may be showing improvement in this area but have not as yet been widely applied.

Finally, the topic of spectral cytometry for biomarkers was revisited. Previous White Papers in Bioanalysis have supported the use of this platform for translational research for biomarker discovery, deep immunophenotyping, and reverse translational science [32]. Specific recommendations for implementation have included qualifying the system for intended use considering fluorochrome specific standardization on assay-by-assay basis, using automated cocktailing/lyophilized cocktails, and leveraging AI/ML tools. A new discussion topic was defining the appropriate processes for clinical implementation (manufacturing specifications, IQ/OQ/PQ). There was agreement that more experience and discussion with vendors are needed before recommendations can be given. In the meantime, specific PQ protocols can be followed specific to fluorochromes (including new fluorochromes as developed amending partial PQ).

There was additional discussion on instrument standardization approaches for spectral flow cytometry platforms. Currently, the most experience in the group is with the Cytek Biosciences platform with limited experience across other spectral platforms. A case study was presented to illustrate that the use of manufacturer provided QC hard-dye beads as a standardization matrix as typically sufficient to achieve the recommended standardization criteria outlined in the CLSI H62; moderate to high intensity beads should have MFIs of ±7% and cells of interest should produce similar plots, with similar MFIs of ±15% for positive staining), equivalent dim resolutions, minimization of gate adjustment [52]. There was an agreement that more manufacture provided fluorochrome specific standards are needed in support of PQ (particularly for new fluorochromes), and that such fluorochrome specific particles may also allow for even tighter standardization metrics compared to the hard-dye bead approach. It was suggested that users may benefit from a hybrid approach to standardization (using fluorochrome specific particles for critical MFI measures).

For validation, LLOQ assessments should be based on intended use. It was agreed that it is a nice-to-have parameter for measures that are unlikely to ever be near a reasonable LLOQ in actual samples, with a consensus recommendation to discuss requirements and limitations with clinical study leads and biostatistics. In instances where LLOQ is defined and there are concerns about reporting BLQ data, it is possible to use two separate data reporting strategies (exploratory vs. validated data) to help ensure data is treated appropriately in applications such as PD modeling and is qualified to support the intended use of the data. Low endogenous or contrived samples (with scientific explanation of suitability) are examples of specimens usable for LLOQ evaluation [52].

7.4. Cell Therapy & Vaccine Novel Approaches for Method Development/ Validation

It is important to evaluate vaccines induced cell-mediated immunity (CMI), which is the ability of immune cells to recognize specific antigens and contribute to protection from disease caused by pathogens. CMI can be a measure of how broad, long-lasting, and cross-protective the immune responses are. The most common methods to measure cell responses, especially T cell responses, are the IFNγ ELISpot and intracellular staining (ICS), using flow cytometry based methods to detect cytokines inside immune cells. T cell responses are commonly quantified in PBMCs. Prior recommendations for these assays have included the need for bridging studies, critical reagent characterization, and instrument harmonization for large and long duration vaccine trials [32]. For example, antibodies should be properly titrated, dilutional linearity of positive staining cells should be shown, and where PBMC are stimulated, unstimulated matched cells should be subtracted from the signal. Due to the heavy reliance on PBMCs in these assays, the recommended pre-analytical considerations include standardizing collection tubes and isolation methods, minimizing processing time from blood raw to PBMC isolation and cryopreservation, temperature control in sample shipment, and validating loss of cell function over time [32].

Case studies were discussed supporting the use of flow cytometry for the measurement of vaccine biomarkers. One case study provided best practices for ICS validation. Among the most important parameters for flow cytometry-based assays in the case study were the assay range and precision. The assay range needs to be determined with appropriate samples, especially at the lower end. This is particularly important if changes in T cell frequencies need to be quantified (e.g., fold rise over baseline) or if responder rates are determined. Assay precision needs to be considered, as well, when measuring vaccine-induced immune responses to ensure the reported changes are beyond assay variability. Analyzing samples from clinical studies needs to be tightly controlled and standardized, as sample analysis may span a long time period. Leukapheresis samples resulting in billions of PBMCs, can be cryopreserved as hundreds of aliquots and have proved useful to monitor assay performance over time. Cryopreserved PBMC samples can also be used to build proficiency panels that can be utilized to compare ICS assays between different laboratories.

These practices were used in another study that qualified an ICS assay to allow for semi-quantitative evaluation of CD4+ and CD8+ T cell responses and was used to assess CMI in vaccine development for different viral pathogens such as influenza and other respiratory viruses. Alternate approaches to the ICS were also discussed, such as the use of Activation-Induced Marker (AIM) Assays. This immunophenotype allows for T-cell receptor-induced-surface markers such as CD25, CD137 (4-1BB), CD69 or CD134 as measurands in response to vaccination [56].

The expert panel reinforced experiences and qualification/validation recommendations from the case studies. Challenges with CMI and ICS assays include variability from stimulation and timing before analysis. Pre-analytical complexities are key considerations. The validation approach recommended was to validate the panel itself, as suggested in CLSI H62. Timing post stimulation within panel (different measures have different optimal timing) should be assessed along with the use of an unstimulated sample to normalize, and determine the viability cutoff before running assay (e.g., >50% viability). An iterative approach described previously can also be used with clinical samples for further validation. Proficiency panels and QC matierials (e.g., Th1 and Th2 panels) are helpful to enable assay and site comparison for large global studies. Validation harmonization efforts are also ongoing (e.g., 2022 White Paper in Bioanalysis and NIBSC reference reagents).

Other case studies were discussed for the use of flow cytometry in CAR T-cell clinical trials. Application of mass cytometry (i.e., CyTOF) as a primary proteomic profiling technology has recently increased in CAR T-cell clinical trials, both in understanding characteristics of patient starting material, as well as the attributes of the CAR T-cell product itself and their respective correlations with clinical outcomes. In a case study, characterization of a past clinical trial, CRB-402, revealed that markers of memory and senescence could be associated with peak CAR T expansion and the positive or negative response to CAR T therapy, respectively.

An important consideration is that patients in CAR T-cell trials donate a large volume of blood for autologous manufacturing, complex analytics, and special lymphodepletion regimens prior to treatment. Thus, there is an urgent need to develop clinical trial monitoring assays that minimize specimen volume and/or minimize visits to medical offices. A case study discussed the development of a high parameter universal patient centric flow cytometry assay, leveraging recent advances in spectral technology to provide six different clinical trial applications from a single tube of blood, namely, determine CAR T cellular kinetics, understand manufacturing success and phenotype maintenance, evaluate tumor burden, and identify predictors of efficacy, safety, and relapse.

A discussion topic for CAR T-cell therapy was how to characterize CAR T-cell manufacturing assays. There was agreement that in trying to predict and understand CAR T-cell therapy success, assays are becoming increasingly complex (an example was given of ∼20 release criteria). LLOQ is often helpful on main readout(s). Quality of reagents and good QC material is critical. Many measures are exploratory for product characterization in manufacturing and post administration, but it is still important to develop good assays and appropriately validate them.

8. RECOMMENDATIONS

Below is a summary of the recommendations made during the 17^th WRIB:

8.1. CBA Novel Strategies for Method Development/Validation

• Prior recommendations were supported for modification of flow cytometry panels for GxP use.

  • Removal of markers should be evaluated with a risk-based approach, and comparability should be shown (e.g., important to demonstrate no effect of removal on gating, and no impact on retained reportables, consider whether there is a risk for a change in stability).

  • Adding a new marker is likely a new assay (significant change) requiring full validation in most instances. Can still apply FFP approach to what “full” validation and comparability testing is needed. QCs should span the range of the assay where needed to demonstrate comparability.

    • In some cases (e.g., testing clones on same fluorochrome), may have method development data to support whether the change is a significant change (full validation) or a non-significant change decision (non-impactful change demonstration).

  • For changes in data analysis (e.g., gating strategy), it is acceptable to use existing validation data. Non-validation data could be used first to adjust analysis to preserve the integrity of the validation dataset as suitable to demonstrate the revised performance.

• It is possible to test clinical samples received outside method validated window. Documentation as imprecise analyte (e.g. tested outside validated stability) along with protection from data misuse is needed.

• Automated gating and unsupervised automated data analysis is showing utility with large clinical datasets.

  • Saves time, reduces analyst number in large trials.

  • Perform software validation, and demonstrate assay is FFP.

  • Experience with automated analysis is still limited; more robustness and reproducibility data are needed.

8.2. Quality Data Generation in High Dimensional Cytometry

• Regarding validation and clinical reporting of AI/ML analysis, most often, AI/ML approaches are used for signal finding followed by manual confirmation through conventional gating.

  • Utilize approaches used in genomics and digital pathology.

  • Retain data traceability.

  • Segregate training/association discovery datasets and testing/confirmation datasets. This requires large data sets and investment.

• It is difficult to compare high dimensional panels across sponsors and trials without harmonization. Increases in dimensionality creates more value in harmonization across assays, but also makes it more difficult.

8.3. Biomarkers Novel Approaches for Method Development/Validation

• The selection for implementation of indirect/competitive and/or direct/bound RO assays usually depends on reagent availability.

  • Prior recommendations were supported for development and validation of these assays like conventional flow cytometry.

• Assays remain most commonly applied in blood. Membrane target expression on cells from dissociated tissues by flow cytometry remains challenging, as it is difficult to demonstrate control and lack of impact from the tissue dissociation process and there are usually other pre-analytical considerations (e.g. biopsy heterogeneity).

• Despite case studies showing successful use of conjugated beads to measure EVs by flow cytometry, more documented experience is needed for widespread use.

  • Size remains challenging, increasingly, flow cytometers (e.g., Attune^™ and CytoFLEX) appear to have advantages over some other platforms in this respect.

  • Standardized isolation procedures for EVs should be identified.

• More experience is needed for clinical implementation of spectral flow cytometry in regard to establishing best practices for instrument qualification.

  • In the meantime, use specific PQ protocols specific to fluorochromes.

• For spectral flow cytometry platforms, manufacture provided QC hard-dye beads are typically sufficient as standardization material but more fluorochrome specific standards are needed.

• For measurements that will be well above reasonable limits of the cytometer, parameters such as LOD and LLOQ can usually be justified as nice-to-have measures based on fit-for-purpose criteria.

8.4. Cell Therapy & Vaccine Novel Approaches for Method Development/ Validation

• ICS by cytometry approaches offers the advantage over ELISPOT approaches in terms of ease of generating cell specific information of cytokines combining with cell lineage and cell activation markers. Challenges include significant variability from cell stimulation and timing before analysis. Pre-analytical complexities are key to address. Current best practices include:

  • Validate panel itself as suggested in CLSI H62.

  • Timing post stimulation within panel (different measures have different optimal timing).

  • Unstimulated sample to normalize and establish critical parameters, such as viability cutoff (e.g., >50% viability).

  • Use clinical samples for further validation as warranted.

  • Proficiency panels (e.g., Th1 and Th2) are helpful to compare assays.

• Flow cytometry is increasingly used to understand CAR T success (manufacturing treatment process and persistence).

  • Assays are taking advantage of technology improvements and becoming increasingly complex.

  • LLOQ is often helpful on main readout(s) as most therapies will approach or surpass the analytic LLOQ for some of these measures at some point in a study.

  • Quality of reagents and good QC material is critical but complex to obtain.

  • Many measures are exploratory for product characterization in manufacturing and post administration.

  • Develop good assays and appropriately validate them.

9. SECTION 3 – Ligand-Binding/Enzyme Assays & Critical Reagents

Isabelle Cludts^d, Mark Dysinger^aj, Uma Kavita^ak, Hiroshi Sugimoto^e, Shannon Chilewski^b, Christine Grimaldi^s, Yong Jiang^y, John Kamerud^ab, Susana Liu^al, Carolina Owen^ah, Nisha Palackal^s, Corinne Petit-frere^am, Samuel Pine^an, Mohsen Rajabi Abhari^g, Kara Scheibner^g, LaKenya Williams^b, Tao Xu^ao, Guodong Zhang^ap

Authors are presented in alphabetical order of their last name, with the exception of the first 4 authors who were session chairs, working dinner facilitators and/or major contributors.

Author affiliations can be found at the beginning of the article.

10. HOT TOPICS & CONSOLIDATED QUESTIONS COLLECTED FROM THE GLOBAL BIOANALYTICAL COMMUNITY

The topics detailed below were considered as the most relevant “hot topics” based on feedback collected from the 16^th WRIB attendees. They were reviewed and consolidated by globally recognized opinion leaders before being submitted for discussion during the 17^th WRIB. The background on each issue, discussions, consensus, and conclusions are in the next section and a summary of the key recommendations is provided in the final section of this manuscript.

10.1. Emerging & Multiplexing Technologies in Bioanalysis

How could the protein quantification methods using Tape Strips be normalized? How do the high-plex proteomics technology platforms (Olink^® vs Somalogics^®) compare? What is the best approach to prepare QC samples for large custom panels? When certain technologies yield results that are significantly higher than results from other platforms, should those technologies be limited to relative assessments? What is the strategy for labelling antibodies or other proteins on paramagnetic beads as well as critical reagent bridging for the homebrew assay using Simoa?

10.2. LBA Tissue Analysis

For bioanalysis in tissues, how much value does comparability of results between platforms (LBA vs. LC/MS) provide?

10.3. Advanced Labeled Critical Reagents Strategies & Hybridoma vs. Phage Display

What are the challenges in labeling reagents for ADA assays for AAV-based gene therapies? How to characterize labeled molecules and evaluate the conjugation efficiency? What are the advantages and disadvantages of Anti-idiotype (anti-ID) antibody generation via hybridoma vs phage display technologies? Critical reagents: retest/expiry extension– any NEW recommendations? Is total concentration the way we should be analyzing recombinant proteins for calibrators or is epitope specific active concentration more relevant? If epitope specific concentration is used, how is this implemented (e.g., how is this shown on a CoA)?

10.4. Advancements in Enzyme Assays

How do we manage critical reagents such as enzymes, substrates, vendor changes, expiration dates, and lot-to-lot variability? What are the considerations for the specific parameters for fit-for-purpose enzyme activity assay validation?

10.5. Novel Modalities Method Development/ Validation Challenges

What are the recommended platforms for evaluating the integrity and/or cleavage of multi-domain masked biologics? What platforms are recommended when these come with non-protein components and payloads? Which established or emerging technologies are valuable for measuring intact vs. fragments of novel (multi-domain) biotherapeutic modalities? For complex novel biotherapeutic modalities with multiple domains, which domains should be measured for PK assays? Are multiple assays needed? Should there be a primary assay and an orthogonal assay in place? What are the recommendations for measuring active exposure of (unmasked) multi-domain biologics? For ADA assessment of novel (multi-domain) modalities, should the standard approach be to set up separate confirmatory tiers to assess domain specificity or should direct (non-confirmatory) ADA binding assays be considered? Is there a risk of false positives or false negatives in the domain specificity confirmatory tiers? What are thoughts around the use of the same lot of endogenous QC samples from a large-scale preparation or bulk pools of endogenous samples in LBA protein biomarker assays to maintain quality of the PD data across different runs? Is this recommended? Are there unforeseen challenges to using the bulk pools approach? Will this suffice to maintain data quality through changes in reference material and reagent Ab lots? Are there general recommendations for the re-purposing of commercial human assay kits for NHP protein biomarker assays? What are the challenges around mixing components from different kits? What are some solutions to the problem of low sensitivity hELISA RNAi PK assays?

10.6. Single Well Analysis (Singlicate) for ADA Assays

What supporting data is available and what are the recommendations for use of singlicate analysis for LBA (both PK and ADA) assays? What are the considerations for LBA methods using singlicate results instead of duplicate?

11. DISCUSSIONS, CONSENSUS & CONCLUSIONS

11.1. Emerging & Multiplexing Technologies in Bioanalysis

Ligand binding assays (LBAs) are used to measure biomolecules such as antibodies and other large molecules. Multiplex LBA technologies add flexibility to sample analyses, but some lack the sensitivity to measure low-level biomarkers or introduce other challenges such as specificity or interference. Newer multiplex platforms discussed were Olink^® (Olink Proteomics; Uppsala, Sweeden) and Somalogic^® (SomaLogic, Inc.; Boulder, CO) which were compared to current standards like MSD and Luminex.

One emerging technology is Olink, whose proprietary Proximity Extension Assay (PEA) technology addresses these issues by using a dual recognition approach with matched pairs of antibodies labeled with complementary DNA oligos and a readout using either quantitative real-time PCR (qPCR) or Next Generation Sequencing (NGS) [57]. The higher-plex products measure the relative abundance of proteins and look for similar trends within the data sets, which can be useful for biomarker screening. The lower-plex kits offer absolute quantification, via calibrator curves, in pg/mL levels of protein. Advantages of the platform include the need for only 1 μL sample/per panel and potential quantification below pg/mL levels, even in samples such as needle biopsies or micro dialysis. There is an option for customization via made-to-measure biomarkers panels. Drawbacks of the platform include the proprietary single-source kits, instruments, and reagents, as well as the training paradigm which is quite extensive. White Papers have shown that this technology generates data comparable with more traditional platforms such as Luminex, MSD, and Mass Spec [58]. Specifically, this comparison also demonstrated parallelism and sensitivity on par with those observed in established platforms such as MSD, and showed improved precision compared to Luminex.

The expert panel also discussed the utility of Olink in comparison to SomaScan, another proteomics technology which uses synthetic reagents called aptamers [59,60]. There was agreement that both Olink and Somalogic are particularly disruptive technologies in highplex proteomics and increase the throughput, accuracy, and selectivity of traditional antibody-based assays that have a limitation in scale. It was recommended that more data from future case studies should be shared, but it was agreed that Olink may be more cost effective between the two for the semi-targeted approach with better selectivity, sensitivity, and lower sample volume requirement.

Another emerging platform, Single Molecule Array (Simoa) from Quanterix^®, is a technology that enhances the sensitivity of ligand binding assays for biomarker analysis. This technology separates and detects single molecules in microwells, allowing accurate and ultra-sensitive measurements of target molecules. A case study of the validation of a PK assay for a monoclonal antibody drug using Simoa was discussed.

This case study spurred discussion of practical considerations for Simoa assay development such as strategies for labelling antibodies on paramagnetic beads and critical reagent bridging for homebrew assays. There was consensus that reagent characterization (i.e., degree of labeling, purity, free label, label efficiency etc.) is critical to maximize the ultra-sensitivity of Simoa. The optimization of the assay buffer for the labeling process may increase its efficiency. The use of 1-Ethyl-3-(3-(dimethylamino)propyl)carbodiimide hydrochloride (EDC) and N-hydroxysulfosuccinimide (sulfo-NHS) in the labeling buffer was recommended.

Overviews of emerging LBA multiplex platforms led to discussion of case studies comparing their performance for clinical trial samples assessment. In one case, 31 samples were tested using commercial biomarker kits for Luminex, Simoa, MSD, Olink, Quansys, and RayBiotech. MSD returned quantifiable results for 68% -100% of the samples for most proteins. This compared to 93%-100% for Olink, 6%-100% for Quansys, and 76%-100% for RayBiotech. The Olink Target 48 panel was identified as an appropriate platform to measure the majority of the biomarkers selected for this comparison work. While MSD and Quansys were listed as viable options for IL-15, IL-6, and TNF-α, Olink is also suitable. IL-2 was only measurable using the Quansys platform. For biomarkers that have no multiplex option identified, additional analysis can be conducted on the Quanterix SP-X.

A novel method for analyzing skin biomarkers was also discussed. Tape stripping is a non-invasive sampling method of the stratum corneum (SC), the outermost layer of the epidermis, allowing further study of SC biomarkers using for example proteomics platforms [61–63]. Tape strips are adhesive collection tools which can be applied consecutively and directly on lesional and normal skin for 3–5 seconds in between applications. The current gold standard for sampling skin is skin biopsies which collect the epidermis and dermis but are invasive and especially difficult to obtain in pediatric patients and/or when recurrent monitoring is required. Sebutape is a unique tape strip type used for measurement of lipids associated with acne while Tape Strips (D-Squame) is widely used for transcriptomics and proteomics in relation to atopic dermatitis and other skin diseases. Tape strip extracts from different dermatoses can be further analyzed through ELISA, western blotting, RNAseq, mass spectrometry, MSD, and high-scale proteomic platforms like Olink, among others. Tape strips have successfully differentiated inflammatory cytokines in normal and atopic dermatitis skin of lesional and non-lesional atopic dermatitis or psoriasis patients [64]. A common hurdle for many non-traditional matrices like these is how to normalize a quantitative analyte.

Regardless of platform, multiplex assays for large panels share common challenges in development and validation, particularly due to reliance on commercial kits and/or vendor provided reagents for home-brew assays (labeling reagents, critical reagents, matrices and buffers). Recommendations were previously provided in the 2022 White Paper in Bioanalysis which are also applicable here [32]. For example, it was agreed upon that multiplex assays are primarily useful in early discovery to identify protein signatures and PD biomarker patterns for hypotheses generation. Biomarkers of interest can then be pursued further in specific assays for [32]. Practical considerations for multiplex panel implementation and validation were discussed. Multiplex technologies should be limited to relative assessments. Industry experts agreed that it is important to distinguish between semi quantitative and quantitative analysis for high plex proteomics technology. It is unpractical to prepare QC samples for thousands of protein analytes. It was therefore recommended to start with system suitability and assay controls for the entirepanel followed by use of additional QC samples for critical decision-making protein biomarker analysis. Each assay platform may use a different mechanism of detection. As such, it is important to confirm assay validity with an orthogonal method. Critical reagent validation also plays a crucial role in the assay orthogonality and robustness and is discussed below.

Recommendations on normalization methods for tape strips were discussed including the use of total protein quantification with BSA standard curve for proteomic studies, which has already been published [62]. There was an agreement that tape strip normalization should utilize methods associated with traditional tissue bioanalysis, use of empty tape strips, tissue biopsies using top layer tissue only, and whole tissue biopsies may be required as assay controls and for assay performance comparison. The inclusion of a pool of tape strip extracts with quantitated levels of keratins or other skin endogenous biomarkers (i.e., microbiome and cytokines were recommended to assess for sample quality. The data can be normalized to total protein content for further comparison across different proteomic studies [32].

11.2. LBA Tissue Analysis

LCMS and hybrid LCMS are specific and robust platforms for tissue protein quantification, especially with direct digestion methods. However, LCMS has drawbacks such as low throughput, limited availability and expertise, and sensitivity challenges that may require IA enrichment. Tissue analysis by mass spectrometry has been previously discussed in White Papers in Bioanalysis from 2017–2022 and the Decennial Index [22]. Key recommended considerations for development of these assays are LCMS method setup and peptide response optimization, extraction and homogenization optimization, antibody screening, surrogate matrix selection, and qualification of parallelism, recovery, and reproducibility.

LBAs are an alternative approach for tissue bioanalysis that is specific, high throughput, and economical, with good sensitivity when using quality reagents. However, LBAs may face issues such as matrix interference, calibrator and QC production, lack of specific antibody reagents, blood contamination, and sensitivity requirements in tissue homogenates.

Considerations for LBA tissue analysis were discussed based on case studies. Many of these considerations are shared with LCMS tissue analysis. This includes optimization of sample processing/homogenization to break tissue structure to form a suspension of tissue solids, proteins, and fluids. Options include mechanical disruption (mortar and pestle, sonication, etc.), number of pulverization cycles under temperature-controlled conditions and use of protein and phosphatase inhibitors and tissue digestion enzymes (Proteinase K, collagenases). It is important to determine if extraction or processing procedures cause loss in analyte recovery from the original matrix. The extraction method determines recovery, accuracy and inter-sample variability. Processing can also affect stability; it is important to determine is the stability of the drug or biomarker in tissue matrix and if it varies by tissue type. Use of protease inhibitors also needs to be assessed as they might not be recommended in some LCMS procedures (e.g. Phosphoprotein analyses).

Another consideration is normalization for different size/weight tissue and total protein content. The options include protein concentration via commercial kits, tissue weight, and tissue housekeeping proteins. These methods have limitations in that they assume target concentration is uniform across tissue and they cannot account for compartment-specific distribution.

Specific considerations for tissue bioanalysis by LBA include whether to use an endogenous or recombinant standard, the need to show parallelism, and use of immunodepletion of blank matrix and dilutional linearity to show endogenous signal is a true positive. Regarding surrogate matrix selection, options may be limited for assay development, and the use of synthetic or surrogate matrices to mimic the matrix may be difficult. Other surrogate matrix options include the use of buffer, serum/plasma, or urine as applicable.

Case studies showed successful LBA quantitation of a mAb drug and a protein target in two different tissue types. Mouse muscle was assayed in one study using TER1 lysis buffer, total protein normalization, and an MSD LBA assay. Another study in rat kidney used cryo-pulverization and TER1 buffer for extraction, tissue weight normalization, and a Gyros/ Singulex Erenna platform for enhanced sensitivity.

There was wide agreement that the decision to use LBA or LC/MS platforms should be made on a case-by-case basis, depending on access to the different technologies, sample volume, assay cost, throughput, project timelines, and sensitivity and selectivity requirements, and ability to manage potential matrix interferences. In general, LBA has a higher throughput and sensitivity, although LC/MS may more easily provide domain specificity. Given the inherent qualitative differences between platforms, coupled with the challenging nature of tissue matrices, there was also consensus that there is little value in showing comparability of results between the two platforms.

11.3. Advanced Labeled Critical Reagents Strategies & Hybridoma Phage Display

Critical Reagents for LBA have been under continual discussion, but there is additional interest in advanced critical reagents for complex AAV based gene therapies that are multi-molecular in nature, such as empty capsids or peptide pools. In 2022, it was recommended that the labeling ratio of ADA assay critical reagents go as low as possible without compromising assay sensitivity. Characterization of reagents post labelling along with the use of BLI and SPR have been used to characterize LBA critical reagents [32].

The physical complexity of AAVs pose significant challenges to their manufacture and characterization with a capsid composed of 60 subunits assembled from 3 proteins. A host of analytical methods were discussed to characterize and monitor unique quality attributes of AAVs including traditional and capillary electrophoresis, western blotting to evaluate viral protein (VP) protein heterogeneity, and ADA assays. These assays bring challenges for ADA labeling. Labeling techniques using biotin and ruthenium to aid in bioanalytical ADA assay creation were discussed along with obstacles encountered during this process, such as characterizing the labeled molecule and evaluating the conjugation effectiveness. Biotinylated AAV at high concentrations produced good signal to noise ratio in the assay. Ruthenium labeled AAV was generated using high challenge ratios; it was found that significantly higher challenge ratios than usually used for mAb drugs are needed to allow development of assays with adequate sensitivity. It was noted that the direct assay format produced more non-specific signals in the serum samples.

These case studies led to recommendations for labelling of reagents for AAV ADA assays. There was agreement that limited information/protocols are available at present, and more data on optimization is needed (e.g., challenge ratio, conjugation conditions, minimum degree of labeling (DoL)). Current advice is to characterize the labeled reagents (capsids) for DoL and for label efficiency as well as possible. It was acknowledged that more case studies and data are needed as there is a gap in extensive characterization of labeled reagents used in AAV ADA assays.

The next critical reagent discussed was the development of anti-idiotype (ID) monoclonal antibodies generated using either hybridoma or phase display. For the development of total pharmacokinetic (PK) assays and total target engagement (TE) assays during pre-clinical and clinical development for therapeutic mAb targeting soluble target, non-neutralizing anti-ID antibodies against a therapeutic antibody (mAb) are key reagents.

There was agreement that hybridomas have a high success rate for generating non-neutralizing anti-idiotype antibodies, as evidenced by the case study presented at the meeting and others. Case studies highlighted that a high-quality antigen (Fab'2 fragment of therapeutic antibody), comprehensive screening and characterization, and high-quality target protein are key to a successful anti-IDs generation campaign. In particular, comprehensive screening and characterization require KD determination, binding profiling to confirm non-inhibition of anti-ID of therapeutic antibody binding to its target protein, matrix interference testing, and pairing for all candidate anti-IDs to select the best non-neutralizing anti-IDs (the clones with the best binding affinity and minimum matrix interference) [65].

Guidance was provided on the advantages and disadvantages for anti-ID generation strategies (Hybridoma vs Phage Display). There was agreement that both methods have advantages and disadvantages. While hybridomas are time-consuming and require the use of animals, consensus was that they are comparatively more effective for generating non-neutralizing anti-IDs and can be produced in-house. On the other hand, phage display is quicker for creating Fab format anti-IDs. However, converting from the Fab format to the full IgG format also takes time. Moreover, it's not the best platform for generating non-neutralizing anti-IDs.

The degree of labeling and label efficiency are crucial for the optimal characterization of critical reagents used in ligand binding assays. Depending on the sample type and configuration, specific LC-MS methods (including the reduced method, intact method with PNGaseF treatment, and Native MS) can provide clear MS profiles. This allows for the calculation of the degree of labeling and label efficiencies. Ultimately, the LC-MS analyses were refined for the characterization of critical reagents, paving the way for the establishment of practices for analyzing similar reagents in the future [66].

Another topic of discussion was the use of epitope specific calibration free concentration analysis (CFCA) to harmonize independent calibrator lots for use in biomarker LBAs. To measure peripheral biomarkers, LBAs require a capture and detection reagent that specifically and selectively recognize complimentary epitopes on a protein of interest. Typically, these assays are calibrated using recombinantly produced surrogate proteins. Lot-to-lot differences in materials can result in bridging studies to re-validate and re-qualify the assay, potentially needing a statistical “fudge factor” to bridge the calibrator lots. Common methods to measure calibrator concentration are UV absorbance, Bradford assay, or Bicinchoninic acid, however, these methods have drawbacks as they measure the total protein in the sample and cannot distinguish between the native, folded protein of interest and misfolded/denatured protein or other contaminants.

The case study proposed an alternative method to total concentration measurement known as epitope specific calibration-free concentration analysis (CFCA) of calibrators and its application in biomarker LBAs. CFCA exploits the well described physical principles of mass transport in a microfluidic chamber, as well as the linear relationship between bound-mass and surface plasmon resonance (SPR) response [67], to define the concentration of an analyte that binds a ligand immobilized on a SPR chip surface. In the context of LBAs, CFCA exclusively reports the concentration of analyte (biomarker) that productively binds a ligand (mAb) and can therefore quantify the concentration of a mAb-specific epitope on a biomarker.

In the case study, four independent recombinantly produced lots of a calibrator were tested to determine the epitope specific active concentration using CFCA after direct conjugation of the capture mAb to the SPR sensor chip. Results showed that the percentage of active biomarker varied lot-to-lot and was considerably lower than the reported total concentration obtained by Bradford assay. When applying the active concentration to calibration curves prepared and tested in the LBA, the results showed a decrease in lot-to-lot variability when compared to total concentration. Other possible uses of the epitope specific CFCA include stability studies to monitor changes in concentration via heat denaturation or deglycosylation. Additionally, this could be used to measure the active concentration of each molecule in a mixed sample, potentially showing a use-case for multiplexed LBA assays.

Recommendations were formed for total concentration vs. epitope specific CFCA for analyzing recombinant proteins used as calibrators for LBAs. There was agreement that total protein is acceptable for use, however CFCA can be employed, keeping in mind that this is specific to the capture and detection antibodies used in the assay. Therefore, it is important to define what the calibrator is being used for. If CFCA is used, it is recommended that it should be specified in the method validation not in the CoA. Additionally, it was recommended to define in SOPs which concentration (total or active) is used.

The final topic of discussion revolved around the practice of re-testing and extending the stability of critical reagents. The consensus was to uphold previous recommendations, which advocate for retesting functionality. There are instances where critical reagents are recertified using biophysical platforms. Specifically, the practice of testing concentration, purity, and binding affinity is deemed acceptable. Furthermore, the expiration date can be reassigned based on reagent's manufacture date and in-house SOP. For instance, a tool antibody, as per in-house SOP, could have an expiration date of 10 years from its manufacture date if stored at -60°C or below.

11.4. Advancements in Enzyme Assays

For AAV-based gene therapy products, the transgene protein is the product of the gene therapy molecule. The enzyme activity of the transgene protein shows the functional effect of the gene therapy molecule after it infects or transduces the cell. Therefore, the connection between the transgene protein expression and its enzyme activity is important for evaluating the biodistribution and pharmacodynamics of the gene therapy product. The development and validation of enzyme assays for this purpose was previously discussed in the 2022 White Paper in Bioanalysis [32]. An enzyme assay measures the enzyme's activity by the amount of substrate used or product made in a reaction. It was recommended previously to use LBA BMV as a general principal with a FFP approach for certain steps for these assays being mindful of potential instability for some analytes/methods.

Additional case studies were presented with focus on the streamlined development and validation strategy of enzymatic activity assays. The method validation for the enzyme iduronate-2-sulphatase (I2S) quantification in serum and method qualification in tissues were performed to support a mouse GLP toxicological study to support developing gene therapy for Hunter Syndrome. Standard curves for I2S quantification ranged from 2.00 to 50.0 μg/mL in serum and 6.25 to 400 ng/mL in the surrogate matrix. Acceptable precision, accuracy, and parallelism in the tissues were demonstrated. To assess the function of the transgene protein, fit-for-purpose method qualification for the I2S enzyme activity in serum was performed, including calibration curve performance, intra- and inter-day precision, selectivity, dilution linearity, short and long-term stability. The observed data indicated that the enzymatic activity in serum increased dose-dependently in the lower I2S concentration range in mice.

The case study led to recommendations and future perspectives for validation of these assays. The fit-for-purpose enzyme activity assay validation may include the linearity of substrate calibration curves, intra- and inter-assay precision, selectivity, dilution linearity, and assay stability. The assay accuracy may be additionally assessed in comparison with the mean value of the QC samples. More data is needed for specific criteria recommendations.

The other consideration for validation was how to manage critical reagents (substrate (i.e., 4-MU), expiration date, lot-to-lot variability). To maintain long-term assay performance, systematic control in critical reagent generation and the bridging of multiple lots is required to assess the robustness of the assay. Securing a large batch of high performing reagent lot may be required. However, if the bridging assay is not feasible, additional work is needed. Extension of the reagent stability by demonstrating assay performance rather than using a new lot of the reagent may be warranted if considerable lot-to-lot variability is observed.

Another critical component of the strategy of enzyme assays is to set “Go and No Go” criteria based on the relationship of transgene and enzyme activity from emerging preclinical data. The clear “Go and No Go” criteria should determine whether the observed transgene protein expression and enzyme activity in the gene therapy treatment is sufficient to achieve a clinically meaningful reduction of the accumulated substrate in humans. As an example, the degree of clinical severity of lysosomal storage disease is often highly correlated with the amount of residual enzyme activity and the level of substrate accumulation [68].

11.5. Novel Modalities Method Development/ Validation Challenges

In addition to the enzyme assays discussed above for AAV gene therapy, challenges in developing and validating ligand binding assays (LBAs) for new modalities, such as CRISPR and oligonucleotide therapeutics (e.g., siRNA quantification for pharmacokinetics or PK), and multi-domain (MDB) recombinant proteins, have been previously explored [32].

Additional case studies were discussed for oligo and multi-domain biologics (MDB) quantification by LBA to update prior recommendations. Integrated bioanalysis approaches were shown for complex MDBs which can undergo biotransformation to activate or inactivate functionality [69,70]. Methods have included indirect quantification of cleaved multi-domain protein therapeutic by LBA, bottom-up IC-LC/MS for masked mAb, and orthogonal semi-quantitative approaches of quantitation & characterization for cleavage events.

LBAs are used for another specialized class of therapeutic, the masked MDBs. Agonistic mAbs and cytokines can activate tumor-fighting T cells but can be too toxic for safe use. “Masked” or “stealth” MDB are a novel approach that combines a targeting (mAb) and an agonist (e.g. cytokine) prodrug domains [71] with a masking/activating moiety that is selectively cleaved by tumor proteases. The agonist is hidden and may be guided by tumor peptides or by a mAb. The agonist is only active in the tumor, only allowing immune activation at the site of action. Masked drugs are complex, and specific PK, PD, and immunogenicity assays should be explored. Molecule stability and immunogenicity risks should also be considered. In the absence of strong evidence of the contrary, it is advisable to assume a higher risk for immunogenicity and to perform a risk assessment to understand potential impact and develop an appropriate immunogenicity strategy. Critical reagents should detect both the mAb and the agonist, and ideally both masked and unmasked molecules. While it may not be necessary to monitor for fragments in in vivo studies, this could aid in the understanding of PK/PD results and provide insights for improving stability.

The case studies led to discussion of which platforms are recommended for evaluating cleavage of multi-domain masked biologics, non-protein components and payloads, and intact vs. fragments of novel (multi-domain) biotherapeutic modalities. Integrated approaches and progressive measurements were recommended. SDS-PAGE and LC/MS analysis may be used to understand the proportion of double unmasked, single unmasked or fully masked forms. LBA can be used to detect unmasked or masked domains if highly specific reagents can be generated. LC/MS may be used to identify cleaved masking peptides, though the value of such data is uncertain. Activity assays may have to be used to further evaluate unmasking and domain integrity.

For MDB PK assays specifically and the need to measure multiple domains with multiple assays, there was updated consensus that if all domains are functional, then PK assays should be set up for all domains. Multiple assays may be needed if the domains differ in content/type of target engaged and function (for contrast, if both domains engage the same target, then a single assay would suffice). Activity assays may not be required for all domains. Progressive measurements and considerations related to the development stage of the drug were recommended. In general, it is acceptable to have multiple assays in the pre-clinical stage, however it is preferable if the assay number can be condensed in the clinical stage.

There was also a discussion on the need to measure active exposure for unmasked MDBs. Measuring “active exposure” at the target site of action may be challenging unless the specific critical reagent to distinguish the active and nonactive biotherapeutics is available for either LBA or LCMS. Accessibility to the analytes of interest and the appropriate matrix is an important consideration (there is perhaps a higher feasibility of acquiring the needed matrix and tissue in pre-clinical studies compared to clinical studies). More discussion with relevant toxicology and clinical teams as well as HA stakeholders may be needed for specific programs. PD markers for efficacy and safety will be even more important with complex multi-domain biologics when faced with PK assessment challenges.

The masked cytokine MDB case study also led to discussion regarding ADA assessment of novel (multi-domain) modalities. The discussion specifically centered on whether the standard approach should be to set up separate confirmatory tiers to assess domain specificity or whether direct (non-confirmatory) ADA binding assays should be considered. There may be a risk of false positives or false negatives in the domain specificity confirmatory tiers. ADA data will continue to be important for safety/efficacy evaluations for muti-domain biologics including the determination of which domains the ADAs bind to. This requires careful consideration of the ADA assay formats. Two separate assays each using a single domain of the multi-domain biologic is one possible format that was recommended. A titer tier may be important to relate the magnitude of ADA with PK/PD effects. There was agreement that further ADA data and discussion is needed.

Finally, there is an emerging need for powerful bioanalytical tools to support ADME or PK/PD of siRNA molecules in circulating plasma and target tissues. Case studies were presented demonstrating the implementation of LBA and hybridization ELISA (hELISA) assays in support of pre-clinical and clinical programs for the determination of siRNA and PD biomarkers in biological samples. hELISA was shown as a method that is designed using capture and detection probes that specifically anneal to the target siRNA and its major metabolites.

There was discussion of solutions to the problem of low sensitivity hELISA RNAi PK assays in comparison to qPCR. There was agreement that sensitivity of hELISA RNAi PK assays is enough for most of the study requirements and it demonstrates comparable sensitivity with qPCR platform based on experience. If not, amplification may help with the sensitivity (either PCR or use of branched stem probes in ELISA). Both hELISA and qPCR have the advantage of quick method development, less complex sample preparation process, and higher throughput. Specifically, using stem probes in the hELISA and qPCR can reach approximately 10 pg/mL to 10 fg/mL of sensitivity. It was recommended that hELISA and qPCR be applied at the early drug discovery stage, and LC/MS be applied at the later stage, especially when active metabolites are more likely to be present [72].

The expert panel also discussed general considerations for novel LBA assays. One scenario where there is a need for the use of large scale or bulk pools of endogenous QC samples is in LBA protein biomarker assays to maintain longitudinal quality of data. There was agreement that the bulk pool can be used for the whole study as long as the analyte is stable, and the bulk prep is stored frozen in single use vials. The large pool that is selected should represent the study population. This is with reference to 100% endogenous QC samples, and not individual QC levels. There was continued recommendation to use QC samples in each batch run. The pooled endogenous samples should be sufficient to run through the whole study. Commercially available matrices are to be considered when patient matrix is limited. Endogenous QCs are also necessary to monitor the assay reproducibility between different runs through changes in reference material and reagent Ab lots.

The final question was for recommendations for the re-purposing of commercial human assay kits for NHP protein biomarker assays and mixing components from different kits. Here, prior recommendations were supported. Homebrew assays may be used, for example, by mixing capture and detection antibodies from different vendors and prior feasibility and qualification. This should be considered on a case-by-case basis. For some target proteins, human kits may not work for NHP if there is no reactivity of the reagents or affinity is poor. It is crucial to further characterize and validate reagents in the kits and not to rely on vendor claims of reactivity. It was also suggested to minimize variability with home-brew kits by minimizing lot changes of components (preparing reagents in bulk with the same lot when possible).

11.6. Single Well Analysis (Singlicate) for ADA Assays

Singlicate analysis for LBA assays has been discussed, as well as their use in PK and ADA assays. In the 2012 White Paper in Bioanalysis, recommendation was made that forgoing singlicate well analysis can be scientifically justified on a case-by-case basis [5]. According to the ICH M10 guidance published in 2022, single-well analysis is acceptable for PK bioanalysis [73] and acceptance criteria does not depend on the number of wells per sample. If the method is validated and statistically supported with single-well analysis, it can be used following published recommendations [74]. However, the bioanalysis of biomarkers and of ADA is not within the scope of ICH M10.

A stepwise, data driven approach was discussed to evaluate the feasibility of singlicate LBA for PK and ADA method development, qualification, and validation to support clinical sample analysis. This approach included 12 Precision and Accuracy runs with 6 runs with duplicate verse 6 runs in singlicate in method validation. The head-on comparison of the \runs, supplementary with statistical analysis result to confirm overall variability of the assay was not impacted by singlicate analysis, and the rest of the validation was completed in singlicate format.

A similar strategic approach was used to evaluate the feasibility of singlicate ADA method development, method qualification, and method validation to support clinical sample analysis. The case study with a multidomain biotherapeutic drug (MBD) demonstrated that a full method validation, including three domain specificities and 2 matrix populations, could be completed with 72 96-well plates, a significant reduction compared to the initial estimate of around 150 96-well plates for a duplicate approach. This represents substantial savings in terms of time and reagents.

These singlicate PK and ADA assays are currently in use to support an ongoing Phase 1 clinical study. The completion of this study is expected in another 15 months, with an estimated total of about 1000 PK and 1000 ADA samples to be analyzed. As of now, approximately, 70% of PK sample analysis had been completed, and 71 ISR samples have been processed. The preliminary ISR results indicate that 90.1% of the samples fall within the acceptance criteria of ±30% from original result. ADA sample analysis is ongoing with the tiered approach and data is pending.

Another case study showed the use of singlicate titer analysis in a phase 3 study with approximately 3000 ADA samples using singlicate titer analysis yielded and significant improvement on cost and efficiency without compromising the data quality and maintaining the familiar expected variable range of ADA titer.

The titer results were in line with results obtained in phase 1 studies which were analyzed in duplicate (low titer, no impact on PK, safety, and efficacy).

These case studies led to recommendations for ADA assays that supported prior recommendations for other LBAs. There was acknowledgement that singlicate analysis is a risk-based approach. Certain analytical platforms have very high levels of precision, making singlicate analysis a viable option on a case-by-case basis. To ensure confidence in singlicate results, at least one method development run was recommended to be performed in duplicate to assess precision and statistical comparison. If a method's intended use is for singlicate sample analysis, it should be validated in singlicate as done in the case of sample analysis. If a method has been validated using duplicates, and/or sample analysis is underway using duplicates, risk assessment of switching to singlicate have to be done, additional validation with proper bridging assessments will be needed. Although singlicate usage has undergone less evaluation for ADA assays, sponsors can perform their own diligence and employ the strategy as appropriate.

12. RECOMMENDATIONS

Below is a summary of the recommendations made during the 17^th WRIB:

12.1. Emerging & Multiplexing Technologies in Bioanalysis

• Olink and Somalogic are particularly disrupting technologies in high-plex proteomics, increasing the throughput, accuracy, and selectivity of traditional antibody-based assays when sample volume is low and there is a need to find protein signatures associated to disease mechanism, druggable targets, and PD effects among others.

  • Olink may be more cost effective for the semi-targeted panels approach with better selectivity, sensitivity, and less sample volume requirement.

• Tape strips are a promising non-invasive sampling technology for stratum corneum biomarkers for use in LBA, LCMS, proteomics, RNAseq, etc.

  • For normalization, utilize methods used in tissue bioanalysis. Total protein content is recommended for protein-based assays.

  • Results may need to be compared to biopsy tissue; and empty strips may need to be used as controls. Keratin or other endogenous skin biomarkers can be quantified for tape strip integrity and method normalization.

• Bead-based antibody platform Single Molecule Array (Simoa) HD-X can be significantly more sensitive than previously developed ELISA and electrochemiluminescence immunoassays.

  • Reagent characterization (i.e., degree of labeling, purity, free label, label efficiency etc.) is critical to maximize ultra-sensitivity of Simoa.

• It is important to distinguish between the semi-quantitative vs. quantitative analysis for high-plex proteomics technologies. For QCs, start with the system suitability and assay controls for the large panel followed by applying QC samples for specific decision-making quantitative protein biomarker analysis.

12.2. LBA Tissue Analysis

• The decision of whether to use LBA or LCMS for tissue analysis should be made on a case-by-case basis and be determined by the needs of the project (cost, sample volume, timelines, use of the data, etc.)

  • In general, LBA has a higher throughput and sensitivity, although LCMS may provide domain specific data.

  • Given the inherent qualitative differences between platforms, coupled with the challenging nature of tissue matrices, it was not recommended to show comparability of results between platforms.

• General considerations for tissue bioanalysis have been discussed extensively in the literature and previous White Papers in Bioanalysis and apply to LBA tissue bioanalysis.

  • This includes response optimization, extraction and homogenization optimization, antibody screening, surrogate matrix selection, and qualification of parallelism.

12.3. Advanced Labeled Critical Reagents Strategies & Hybridoma/Phage Display

• Limited information/protocols are available for labelling AAV ADA assay reagents (e.g., challenge ratio, conjugation conditions, minimum DoL needed). High challenge ratios may help increase assay sensitivity, but more case studies and data are needed before recommendations can be made.

  • In the interim, it is important to characterize the labeled reagents (capsids) for DoL and for labeling efficiency.

• Multiple methods are available for anti-drug non-neutralizing anti-ID reagent generation.

  • Hybridomas are time intensive but can be made in-house.

  • Phase display is faster and can generate neutralizing anti-IDs but is not as accessible.

• Recombinant protein epitope specific active concentration determination by SPR is an innovative alternative to total protein concentration determination.

  • The need for active concentration determination is target specific (e.g., high lot-to-lot variability) and total protein is still typically acceptable.

  • If active concentration is used, specify it in the method validation and SOPs (not CoA).

• Prior recommendations were supported for retesting and extending the critical reagent expiry date.

  • Functionality must be retested; biophysical platforms can be used for binding affinity determination along with testing concentration and purity.

12.4. Advancements in Enzyme Assays

• The fit-for-purpose enzyme activity assay validation may include the linearity of substrate calibration curves, intra and inter-assay precision, selectivity, dilution linearity, and assay stability. The assay accuracy may be additionally assessed in comparison with the mean value of the QC samples during the method qualification.

• Setting “go/ no go” criteria is critical for the development of enzyme activity assays for transgene proteins.

  • Assess preclinical data to determine whether the observed transgene protein expression and enzyme activity correlation is sufficient to achieve the clinically meaningful reduction of the accumulated substrate in humans.

12.5. Novel Modalities Method Development/ Validation Challenges

• Integrated approaches and progressive measurements were recommended for MDB cleavage, non-protein components, payloads, and intact vs. fragments of MDBs.

  • SDS-PAGE and LC/MS analysis may be used to understand the proportion of double unmasked, single unmasked or fully masked forms.

  • LBA can be used to detect unmasked or masked domains.

  • LC/MS may be used to identify cleaved masking peptides.

  • Activity assays may be used to further evaluate unmasking and domain integrity.

• Prior recommendations were supported that all functional domains in MDBs must have PK assays.

  • Multiple assays will also be needed if the domains differ in content/type of target engagement and function.

  • Activity assays may not be required for all domains.

• Measuring “active exposure” at the target site of action may be challenging unless specific critical reagents to distinguish the active and nonactive biotherapeutics are available. PD markers for efficacy and safety may be more informative.

• ADA data will continue to be important for safety/efficacy evaluations for muti-domain biologics, especially for masked cytokine MDBs.

  • Separate assays each using a single domain of the multi-domain biologic is one possible format that was discussed.

  • A titer format or S/N may be important to relate the magnitude of ADA with PK/PD effects.

• Regarding the challenge of developing an endogenous QC preparation for biomarker assays, prior recommendations were supported.

  • If the analyte is stable, a bulk prep can be used if stored in single use vials.

  • The large pool that is selected should represent the population and endogenous QC samples should be used in each batch run.

• Prior recommendations were also supported for adapting and mixing commercial biomarker kits for different species.

  • Homebrewed assays may be used case-by-case, for example, by mixing capture and detection antibodies from different vendors. Validate the reagents in the kits and do not rely on vendor claims of cross-reactivity.

12.6. Single Well Analysis (Singlicate) for ADA Assays

• Considerations for the use of singlicate analysis for ADA assays instead of duplicate are similar to those recommended previously for PK assays.

• A risk-based approach is recommended. Certain analytical platforms have very high levels of precision and make singlicate analysis a viable option on a case-by-case basis.

• At least one method development run should be performed in duplicate to assess precision.

• If a method's intended use is for singlicate sample analysis, it should be validated in singlicate. If a method has been validated and run using duplicates, it is feasible to switch to singlicate with proper bridging assessments.

• Singlicate ADA assays have undergone less evaluation than singlicate in PK and biomarker methods, and more data is needed for wider recommendations. Sponsors can perform their own diligence and employ the strategy as appropriate.

Financial disclosure

The authors have no financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.

Acknowledgments

• US FDA, EU EMA, UK MHRA, Austria AGES, Belgium FAMHP, Netherlands IGJ, Brazil ANVISA, Health Canada, Japan MHLW, and WHO for supporting this workshop

• All Session Chairs & Working Dinner Facilitators for chairing the workshops and the White Paper discussions: Dr. Mitra Azadeh (Ultragenyx), Mr. Mike Baratta (Takeda), Dr. Gopa Biswas (US FDA), Dr. Katherine Block (Genentech), Mr. Mark Dysinger (Alexion), Dr. Seongeun Julia Cho (US FDA), Dr. Isabelle Cludts (UK MHRA), Ms. Kelly Coble (Boehringer Ingelheim), Dr. Vilma Decman (GSK), Dr. Steven Eck (AstraZeneca), Dr. Anna Edmison (Health Canada), Dr. Fabio Garofolo (BRI Frontage), Dr. Swati Gupta (AbbVie), Dr. Shawna Hengel (Seattle Genetics), Ms. Sarah Hersey (BMS), Dr. Allena Ji (Chiesi), Dr. Wenying Jian (Janssen), Dr. Surinder Kaur (Genentech), Dr. Uma Kavita (Spark Therapeutics), Dr. Christopher Kochansky (Exelixis) Dr. Yi-Dong Lin (Takeda), Dr. Meena (Stoke), Dr. Johanna Mora (BMS), Dr. Rachel Palmer (Sanofi), Dr. Susan Richards (Sanofi) Dr. John Smeraglia (AstraZeneca), Dr. Ivo Sonderegger (Takeda), Dr. Yuan Song (Genentech), Dr. Hiroshi Sugimoto (Takeda), Dr. Matthew Szapacs (AbbVie), Dr. Martin Ullmann (Fresenius Kabi), Dr. Meenu Wadhwa (UK MHRA), Dr. Jian Wang (Crinetics), Dr. Russell Weiner (Takeda), Dr. Long Yuan (Biogen), Dr. Yiyue (Cynthia) Zhang (US FDA)

• All the workshop attendees and members of the Global Bioanalytical Community who have sent comments and suggestions to the workshop to complete this White Paper

• Future Science Group as a trusted partner

Competing interests disclosure

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, stock ownership or options and expert testimony.