Scientific and Regulatory Considerations for the ... · Bi,o m:ark,er Assay CoUaborativ,e...

U.S. FOOD & DRUG ADMINISTRATION Duke MARGOLIS CENTER for Health Policy I CRITICAL PATH \~ INSTITUTE

Scientific and Regulatory Considerations for the Analytical Validation of Assays

Used in the Qualification of Biomarkers in Biological Matrices

June 14-15, 2017 [email protected]

1

mailto:[email protected]

U.S. FOOD & DRUG

2

SCIENTIFIC AND REGULATORY CONSIDERATIONS FOR THE ANALYTICAL

VALIDATION OF ASSAYS USED IN THE QUALIFICATION OF BIOMARKERS IN

BIOLOGICAL MATRICES JUNE 14, 2017

OPENING REMARKS

ShaAvhrée Buckman-Garner, MD, PhD, FAAP Director, Office of Translational Sciences, CDER, FDA

" ~--~ ••

www.fda.gov

Bi,om:ark,er A ssay CoUab orativ,e Evident i,a1ry Consider,aU ons

Writ ing 1Gr,oup, CriUca1I Path lnstit ut ,e (C-P,ath)

S1iev e 111 P _~ ,eo t e r ic Co111su lt i111g ,& Jo h111 M ic hae l Saiue r, Cr it ical Path l 111stit L1 fie

Co111tr i b ut i 111g A ut ho rs:

Shash i Amur, U.S_ FDA; Ji r i Mb_~ Pfize r; A maindai 5aike r, Cr it irnl P,at h l111sl:it utie; Rob ert Bec ke r, U.S_ FDA; Jenn ife r aYlli§!. Cr it ical P,ait h Inst it ut e; Rob ert De,ain, Eli Li lly; M airt h ai Dono,ghue, U.S_ FDA;

Russe ll Gr;ant, Laib Corp; Sfiev e 111 G!.!tmsn.U~ ~ijJ; Ky li e Haskins, U.S_ FDA; Joh111 !Kiadaiv il, U.S_ FDA; l'l•I I.cholas King, Cr it icai l Path l111stit ut e; Jea111 Le e,~_cm; ~ u.s_ FDA ;~~ ~b{m; Sh elli S:~~.C:, Pfoe r, M ~J~.!l9l~~~ T:akie,dai; Sll e Jaine W ang, U.S_ FDA

THANK YOU

Duke-Margolis C-Path Institute

• Meredith Freed • Martha Brumfield

• Elizabeth Richardson • John Michael Sauer

• Nick King

3

" ~--~ ••

www.fda.gov

WHY ARE WE HERE TODAY?

• Review and refine a draft set of best practices and performance characteristics for biomarker assay validation for the purpose of biomarker qualification

• Using case studies, explore methodological implications of the framework and identify any remaining gaps

• Determine a path forward

4

-

" ~--~ ••

-----www.fda .• -

FDA LAID THE FOUNDATION FOR QUALIFICATION OF DRUG DEVELOPMENT TOOLS

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Histopath

Guidance

(draft)

CPIM

introduced

BMQ

Guidances and MAPPs

FDA-EMA collaboration

CPIM

LOS

PhRMA-

FDA

Workshop

Brookings

Meeting

LOS

(7)

IOM

meeting

Invasive

Aspergillosis BM

HHMI Level

of Evidence

Meeting

LOS

(1)

Brookings

Meeting

LOI Harmonization

CPIM

Guidance

and MAPP

Quarterly

EMA-FDA

teleconferences

M-CERSI

Meeting

FDA-FNIH

Workshop

LOS

(5) LOS

(1)

White Paper

Meeting/workshop

Survey

Guidance DDT

Qualification (draft)

Guidance DDT

Qualification (final)

CDER DDT

Qualification

MAPP

FR notice

BQ survey

OND

survey

1st nephrotox BMs

CP Opportunities

List

2nd

nephrotox

BMs

Cardiac

toxicity BMs

2016 2017

Total Kidney

Volume in

ADPKD

Plasma

fibrinogen in

COPD

5

" ~--~ ••

www.fda.gov

WHAT HAVE WE LEARNED?

6

" ~--~ ••

www.fda.gov

ANALYTICAL AND CLINICAL VALIDATION CONSIDERATIONS IN BIOMARKER QUALIFICATION

The Specific Context of Use for a Biomarker Drives the Extent of

Evidence Needed for Qualification

Analytical Validation Clinical Validation

(establish performance characteristics and (establish that the biomarker acceptably

acceptance criteria) identifies, measures, and supports the COU)

Pre-Analytical Reference Clinical Sampling and Assay Analytical Rigor/ Study Design Benefit/Risk Ranges/ Meaningfulness/ Handling/

Performance Reproducibility Acceptability Assessment Decision Points Decision Points Stability Characteristics

7

" ~--~ ••

www.fda.gov

DEVELOPMENT OF QUALIFIED BIOMARKERS

Biomarker Discovery

and Development

Letter of Intent Qualification

Plan

Full

Qualification

Package

FDA Review

and

Decision

FDA Review

and

Decision

FDA Review

And

Decision

Prospective Studies

Letter of Support

Scientific Publications

Scientific Workshops

Scientific Community Input

Analytical and Clinical Validation

Critical Path Innovation

Meetings

FDA

Publically

Disseminates

Qualification

Determination

Expanded

Use in

Clinical Trials

Qualification

for Context of

Use (COU)

Additional Information to Expand COU

8

www.fda.gov

Drugs

Home > ONgs ) Oe\ielopmen1&Approll'alProcess(Orugs) > OrugDewlopment ToolsOualfficationPrograms

Drug Development Tools Qualification Programs

Animal Model O ua lifx:a tion Program

Clinical Outcome Ass.e.ssment Qualification Program

8iomarker Oitalification Program

Updated Process for Qualification of Drug Development Tools Under New FD&C Act Section 507

f 81-lAR.E 'I r .... 'EE- In LINKEOIPf @ = ',I IT • EMAIL B PRIKT

Under fhe 21 st Century Cures Ac.t1 en.acted on December 13 , 2016 , the new section 507 Qualification o f Drug

Developmen t Tools (ODT-s) we.s added to the Federa.1 Food, Drug. and Cosmetic Act and formally establishes an

updated , multi•stage prooes-s for DOT qualification . Thi s process includes three submiss.ion m ilestones: the Letter o(

Intent (LOI), the Quali ficstion Plan (OP). and the Full Qualification Package (FQPf. Section 507 else, includ es.

transparency provisions that apply to requesters ' submissions and FDA's formal written determination s in response

to such subm issk>ns3. Cons.i.stent w ith the transparency provisions of section 507, FDA intends to publicly post the information containe

Date of receipt and status (under consideration. accepted. 01 declined)

Requestoc Name

DDT Type (e.~. Biomar1cef. Clinical Outcome Assessment)

DDT NameJOescription

Proposed Coo:ext of Use (COU)

DDT Oe'leloper's Submitted LOI Summary

Drug development need the DOT is Sltended to address

Notation if FDA consulted external ~

FDA Formal Witt..n Determination le-."t-21' (accepiling or declining to accep1 the LOI)

Date of receipt and status (imder consideration. accepted. or declined)

Requestoc Name

DDT Type (e.~. 8iomar1cef. Clinical Outcome Assessment)

DDT NameJDescription

https://www.fda.gov/Drugs/DevelopmentApprovalProcess/DrugDevelop mentToolsQualificationProgram/ucm561587.htm

9

https://www.fda.gov/Drugs/DevelopmentApprovalProcess/DrugDevelop

" ~--~ ••

www.fda.gov

WHY IS THE TIMING OF THIS MEETING CRITICAL?

10

" ~--~ ••

www.fda.gov

(1) Congress shollld promote and facilitate a

2 collaborative effort among the biomedical research

3 consortia described i11 subsection (a)(3)-

4 (A) to develop, through a transparent pub-

5 lie process, data standards and scientific ap-

6 proaches to data collection accepted by t he

7 medical and clinical research com1mmity for

8 ptU'poses of qualifying drug development tools;

9 (B) to coordinate efforts toward developing

IO and qualifying drug development tools in key

I I therapeutic areas; and

I 2 (C) to encourage the development of acces-

13 s ible databases for collecting relevant drug de-

14 velopment tool data for such pllrposes; and

I 5 (2) an entity seeking to quali fy a drug develop-

16 ment tool should be encouraged, in addit ion to con-

) 7 sultation with the Secretary, to con. ul t with bio-

l 8 rned ical research consortia and other individuals and

I 9 entities with e.:-..-pert lrnowledge and insights that may

20 assist t he requestor and benefit t he process for uch

2 I qual ification.

21ST CENTURY CURES ACT

11

" ~--~ ••

www.fda.gov

WHAT WOULD CONSTITUTE A SUCCESSFUL WORKSHOP?

•We understand that we – as a research community – “own” this. • “The more of the concepts for necessary analytical validation

that can be incorporated in the earliest experiments, trials, etc. then the more expeditiously we can achieve qualification. It is not only companies who own this or those who undertake biomarker qualification, the entire community, meaning each of us, has a contribution.” Martha Brumfield, C-Path Institute

•We begin to articulate core expectations for the validation of a fluid biomarker assay used in qualification.

•We understand how Benefit/Risk and Context of Use defines additional assay performance requirements/expectations.

•We develop greater understanding of a potential framework that may be applied while considering the unique attributes of biomarker being qualified under a specific COU.

12

" ~--~ ••

www.fda.gov

THANK YOU!

CDER/OTS

• Shashi Amur

• Aloka Chakravarty

• Ru Chen

• Kylie Haskins

• Laurie Muldowney

• Marianne Noone

• Mike Pacanowski

• Ameeta Parekh

• Sarmistha Sanyal

• Sue Jane Wang

CDER/OND • Elizabeth Hausner • Chris Leptak • Aliza Thompson

CDER Janet Woodcock

CDRH • Bob Becker • Dan Kraniak • Vasum Peiris

OC/OPT • Suzie McCune

13





14


-

-

U.S. FOOD & DRUG

FDA EFFORTS TO SUPPORT BIOMARKER DEVELOPMENT AND QUALIFICATION

CHRISTOPHER LEPTAK, M.D., PH.D. DIRECTOR, OND REGULATORY SCIENCE PROGRAM CO DIRECTOR, BIOMARKER QUALIFICATION PROGRAM

June 14 15, 2017

~ !,! __ :;;; ••

www.fda.gov

Disclaimers

• Views expressed in this presentation are those of the speaker and do not necessarily represent an official FDA position

• I do not have any financial disclosures

regarding pharmaceutical drug products

16

!,! __

:;;; ••

Right target

• Strong link between target and disease • Differentiated efficacy

f f • Available and predictive biomarkers ~-----------------~

Right tissue

• Adequate bioavailability and tissue exposure • Definition of PD biomarke rs

..... :.··.·.·.·.;}'.;:'.,'; : G~e;::::e::~:::;~~u~f~~:~l\~;:r~:~~o~: n;cal PK/PD Differentiated and clear safety margins

... {···· ........ Understanding of secondary pharmacology risk : ···· ······························· Understanding of reactive metabolites, genotoxicity, drug-drug interactions

~::~:: ...... :: .. ·..................................... Understanding of target [;ab;[;ty ·····················--····································· ~R_ig_h_t_p_a_t_ie_n_t_s ________________ ~

··············· Identification of the most responsive patient population Definition of risk-benefit for given population

e rr m-anc-et access, paye r ana provrae Personalized health-care strategy, including diagnostic and biomarkers

www.fda.gov

Key Contributors to Drug Development Success

Coo17 Adapted from Cook et al., Nature Reviews Drug Discovery 13:2419-431 (2014)

Feral

Partners

" ~--~ ••

www.fda.gov

COMPONENTS OF DRUG

DEVELOPMENT SUCCESS

Each of these elements share importance to drug approval

Since any element can lead to failure, important

Biomarker

Clinical Trial

Design/ Endpoint

Patient Population

Assay

to optimize as appropriate and feasible

18

~ !,! __ :;;; ••

www.fda.gov

FDA Regulatory Approach to Biomarkers

• Definition: a defined characteristic that is measured as an 1) indicator of normal or pathogenic biological processes or 2) response to an intervention.

• Broadly defined, with multiple biomarker types including molecular, histologic, radiographic, and physiologic. (i.e., serum protein, change in tumor size by imaging study, algorithm for QT determination on ECG)

• Characteristic is not a clinical assessment of how a patient feels, functions, or survives (contrasted with Clinical Outcome Assessments or COAs)

• Although a biomarker may be used by clinical or basic science research communities, regulatory acceptance focuses on a drug development context that is supported by data for that context. Considerations include:

• Reproducibility of data (e.g., high rate of discordant conclusions RE biomarkers in the published literature)

• Adequacy of the analytic device to assess biomarker’s reliability

• Feasibility of the biomarker should a drug be approved (e.g., will the analytic be widely available and capable of integration into clinical practice paradigms)

19

• Biomedical scientists

• Translational and clinical researchers

• Medical product developers

• Patient/disease advocacy groups

• Government officials

" ~--~ •• - - -

-

--------------------------------- www.fda.gov

• Clinicians

BEST: BIOMARKERS, ENDPOINTS,

AND OTHER TOOLS RESOURCE

• A glossary of terminology and uses of biomarkers and endpoints in basic biomedical research, medical product development, and clinical care

• Created by the NIH-FDA Biomarker Working Group

• Publicly available at http://www.ncbi.nlm.nih.gov/books/NBK326791/

• BEST harmonizes terms and definitions and addresses nuances of usage and interpretation among various stakeholders, including:

• Biomedical scientists

• Translational and clinical researchers

• Medical product developers

• Patient/disease advocacy groups

• Government officials

• Clinicians

20

http://www.ncbi.nlm.nih.gov/books/NBK326791/

!,! __

:;;-; ••

www.fda.gov

Biomarker Classes from a Drug Perspective

Susceptibility/Risk: Indicates potential for developing disease before it is clinically apparent

(e.g., BRCA mutations and development of breast cancer)

Diagnostic: 1) Detects presence of a disease or condition or 2) identifies patient subsets

(e.g., HbA1c to aid in diabetes diagnosis)

Monitoring: Assesses disease status, including degree or extent, through serial measurement

(e.g., INR and anti-coagulation status)

Prognostic: Identifies likelihood of a clinical event, disease recurrence, or progression, in in the

absence of a therapeutic intervention (e.g., BRCA mutations and breast cancer recurrence)

Predictive: Identifies patients who are more likely to experience a favorable or unfavorable

effect from a specific treatment (e.g., HLA-B5701 and risk of severe AE with Abacavir)

Pharmacodynamic/Response: Indicates that a biological response has occurred in a patient

who has received a therapeutic intervention. May become a clinical trial endpoint and for a very

small subset, surrogate endpoint. (e.g., sweat chloride and response to CFTR agents)

Safety: Indicates the likelihood, presence, or extent of toxicity to a therapeutic intervention when

measured before or after that intervention (e.g., QTc and Torsades)

21

" !,! __

:;;; ••

~~~---------~~-~--~~-~-----------~

[ l

I ,

I r-------i r----i r-------,

[ ] [ l [ l www.fda.gov

“Fit for Purpose”: BEST BiomarkerClasses in Perspective

Susceptibility/Risk

Pathologic

Changes

Descriptive

Time progression

Key factors / events

Altered

Physiology

Descriptive

Threshold of concern Clinical

Disease

Diagnostic

Monitoring

Prognostic

Change in

Physiology

Non-Progression

Or Reversal

Improved

Clinical Benefit

“Normal”

Physiology

Change

Therapeutic Intervention

Pharmacodynamic

Predictive

Safety

Response Surrogate Endpoint

22

" ~--~ ••

www.fda.gov

BIOMARKER INTEGRATION

INTO DRUG DEVELOPMENT

Biomarker

Qualification

Program

Drug Approval

Process

Scientific

Community

Consensus

Note: These pathways do not exist in isolation and many times parallel efforts are underway within or between pathways. All share common core concepts, are data-driven, and involve regulatory assessment and outcomes based on the available data.

Facilitating Biomarker Development: Strategies for Scientific Communication, Pathway Prioritization, Data-Sharing, and Stakeholder23 Collaboration; Published June 2016, Duke-Margolis Center for Health Policy

" ~--~ ••

.-,,,!!!!

www.fda.gov

DDT QUALIFICATION AT CDER

Guidance for Industry and FDA Staff:

Qualification Process for Drug Development

Tools http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/

Guidances/UCM230597.pdf

Drug Development Tools (DDT) Qualification

Programs Webpage on FDA.gov http://www.fda.gov/Drugs/DevelopmentApprovalProcess/DrugDevelopmentToolsQualific

ationProgram/default.htm

24

http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM417627.pdfhttp://www.fda.gov/Drugs/DevelopmentApprovalProcess/DrugDevelopmentToolsQualificationProgram/default.htm

" ~--~ ••

www.fda.gov

TYPES OF SUBMISSIONS WE ARE SEEING FOR BIOMARKER

25

QUALIFICATION

19% Patient Selection

26% Preclinical Safety

30% Response

22% Clinical Safety

4% Monitoring

N=27

" ~--~ ••

www.fda.gov

LIST OF FDA-QUALIFIED BIOMARKERS

General Area Submitter(s) Biomarker(s) Qualified for Specific

Contexts of Use

Issuance Date with Link

to Specific Guidance

Supporting

Information

Nonclinical

Predictive Safety and Testing

Consortium (PSTC),

Nephrotoxicity Working Group

(NWG)

Urinary biomarkers: Albumin, β2

Microglobulin, Clusterin, Cystatin C,

KIM-1, Total Protein, and Trefoil

Factor-3

4/14/2008: Drug-Induced Nephrotoxicity

Biomarkers Reviews

Nonclinical

International Life Sciences

Institute (ILSI)/Health and

Environmental Sciences Institute

(HESI), Nephrotoxicity Working

Group

Urinary biomarkers: Clusterin, Renal

Papillary Antigen (RPA-1)

9/22/2010: Drug-Induced Nephrotoxicity

Biomarkers Reviews

Nonclinical PJ O’Brien, WJ Reagan, MJ

York, and MC Jacobsen

Serum/plasma biomarkers: Cardiac

Troponins T (cTnT) and I (cTnI)

2/23/2012: Drug-Induced Cardiotoxicity

Biomarkers Reviews

Clinical Mycoses Study Group Serum/bronchoalveolar lavage fluid

biomarker: Galactomannan

10/24/2014: Patient Selection Biomarker for

Enrollment in Invasive Aspergillosis (IA)

Clinical Trials

Reviews

Clinical

Chronic Obstructive Pulmonary

Disease (COPD) Biomarker

Qualification Consortium

(CBQC)

Plasma biomarker: Fibrinogen

7/6/2015; Prognostic Biomarker for

Enrichment of Clinical Trials in

Chronic Obstruction Pulmonary Disease

(COPD)

Reviews

Clinical Polycystic Kidney Disease

Outcomes Consortium

Imaging biomarker: Total Kidney

Volume (TKV)

8/17/2015: Prognostic Biomarker for

Enrichment of Clinical Trials in Autosomal

Dominant Polycystic Kidney Disease

Reviews

www.fda.gov/biomarkerqualificationprogram

26

http://www.fda.gov/biomarkerqualificationprogram

-' !!._ :;;-; ••

1938 Food , Drug & Cosmetic Act (safety)

• I 1940s

-

Timeline of U.S Drug Regulation

1951 1962 1983 1992 2004 Durham- Kefauver-Harris Orphan PDUFA Critical Humphrey Amendment Drug Path Amendment (efficacy) Act Initiative (prescriptions)

• I ••• • I •:, .. , Ill I I

-2007 2008 2010 2014 Biomarker First DDT Draft DDT Final Qualification Biomarker Guidance Guidance Pilot Process Qualification Initiated

Timeline of U.S. Biomarker Regulation

www.fda.gov

Keeping perspective… We have been developing drugs formuch longer than biomarkers

27

" ~--~ ••

www.fda.gov

CONSIDERATIONS FOR BIOMARKER UTILITY

Context of Use (COU): 1) BEST biomarker category and 2) how the

biomarker impacts the clinical trial or drug development program

What question is the biomarker intended to address. Examples include:

o Inclusion/exclusion criteria for prognostic or predictive enrichment?

o Alter treatment allocation based on biomarker status?

o Result in cessation of a patient’s participation in a clinical trial because of safety concern?

o Result in adaptation of the clinical trial design?

o Establish proof of concept for patient population of interest?

o Support clinical dose selection for first in human or Phase 3 studies?

o Evaluate treatment response (e.g. pharmacodynamic effect)?

o Support regulatory acceptability of a surrogate endpoint for accelerated or traditional approval?

“Total Kidney Volume, measured at baseline, is a prognostic enrichment biomarker to select patients with ADPKD at high risk for a progressive decline in renal function (defined as a confirmed 30% decline in the patient’s estimated glomerular filtration rate (eGFR)) for inclusion in interventional clinical trials. This biomarker may be used in combination with the patient’s age and baseline eGFR as an enrichment factor in these trials.”1

1 https://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM458483.pdf 28

https://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM458483.pdf

~ !,! __ :;;; ••

www.fda.gov

The need and strategy for stakeholder

engagement…

• Since the founding of the Biomarker Qualification Program (BQP), FDA and stakeholder communities have continued to gain experience to better understand the level of evidence needed to support qualification of a biomarker.

• The 21st Century Cures legislation has called for the development of guidance(s) outlining an evidentiary framework to support biomarker qualification.

• Dr. Woodcock has emphasized publicly that this not FDA’s responsibility alone (“It takes a village to raise a biomarker”)

• In 2015 and 2016, a number of stakeholder groups, including FDA, convened a series of meetings to start to understand the components of an evidentiary framework.

29

!,! __

:;;; ••

www.fda.gov

Prior workshops

• Evidentiary Considerations for Integration of Biomarkers in Drug Development

- U. of Maryland CERSI/FDA/Critical Path Institute, August 21-22, 2015

• Facilitating Biomarker Development and Qualification - Brookings Institution, October 27, 2015

• Collaboratively Building a Foundation for FDA Biomarker Qualification - National Biomarker Development Alliance, December 14-15, 2015

• BEST (Biomarkers, EndpointS, and other Tools) Glossary - FDA/NIH Joint Leadership Council, NLM

• Biomarker Qualification Workshop: Framework for Defining Evidentiary Criteria

- FNIH Biomarkers Consortium/FDA, April 14-15, 2016

30

" ~--~ •• In Drng Develllopment

INeed

Evaluate Compa red to Status Quo

L)cou ,I r

' ' ' ' ' ' ' ' . ' ' , I :_ _________ ¥~

., Class of Biomarker. •

., What is the question the biomarker i.s • addre.ss.ing.

•

Improved sensitivity

Improved selectivity

Mechanistic context

To Patient

l Informs Required Stringency

of EC

•

•

~

Risk [ '\ / ,I

Consequence of false positive

Consequence of false negative

r

.,

.,

. ,

.,

.,

.,

.,

Evi,dentiary C1riteria

Characterization of Relationship Between the Biomarker and Clinicaf Outcome

Biological Rationale f or Use of Biomarker (if Known)

Type of Data and study Design ,(i.e . Prospective, Retrospective~ etc.)

Independent Data Sets f or Qualffication

Comparison to current standard

Assay performance

Statis tical Meth ods to Use

www.fda.gov

CONCEPTUAL FRAMEWORK FOR BIOMARKER DEVELOPMENT FOR REGULATORY ACCEPTANCE

31 https://fnih.org/what-we-do/biomarkers-consortium/programs/framework-for-safety-biomarkers

!,! __

:;;; ••

www.fda.gov

Thoughts for the day

What are the distinctions between analytical validation for clinical practice use and use in a clinical trial context?

How does a biomarker’s category (BEST) and context of use in drug development impact what analytical performance characteristics are of greatest importance?

What level of certainty in analytical performance is necessary to mitigate the potential benefits and risks to patients in a clinical trial context?

How does the phase of drug development (early vs late) impact that level of certainty needed?

What parallels can we draw between how diagnostics are used in drug-specific clinical trial contexts (under IND) including the level/extent of analytical validation?

32

" ~--~ ••

U.S. FOOD & DRUG ADMINISTRATION

---------------www.fda.gov

THANK YOU FOR YOUR ATTENTION

33





34



Overview of the Analytical Validation White

Paper Objectives and Key Terminology

Steven P. Piccoli, Ph.D. CEO, Neoteric Consulting LLC

35

CRITICAL PATH INSTITUTE

B7z&\L\ RESEARCH • PROGRESS • HOPE

~ FNIH ; Foundation for the National Institutes ol Health

Framework for Defining ~ Evidentiary Criteria ) for Biomarker Qualification

Final Version

EvidentiaryCriteriaWritingGroup

Biomarker Qualification Evidentiary Considerations Activity Map

Biomarker Evidentiary Considerations Framework

Working Group

Biomarker Qualification Assay Validation Expectations

Working Group

Biomarker Qualification Statistical Considerations

Working Group

M-CERSI Evidentiary Consideration Meeting Brookings Institute Meeting FNIH BC Evidentiary Consideration Meeting Biomarker Qualification Statistical Considerations Meeting Biomarker Qualification Assay Validation Expectations Meeting

36White Paper Overview and Terminology

S. P. Piccoli

http://www.phrma.org/

American Association of Pharmaceutical Scientists

Crystal City VI: BMVon Biomarkers

R1naln11 nc• Baltlmort Harborplac1 Hot1I Baltlmor1, Md.

Biomarker Assay Validation Activity Map


S. P. Piccoli

Pertinent FDA Guidances

• 2001 – Bioanalytical Method Validation

• 2007 – In Vitro Diagnostic Multivariate Index Assays [Draft]

• 2013 – Bioanalytical Method Validation [Draft]

• 2014 – Qualification Process for Drug Development Tools

• 2014 – In Vitro Companion Diagnostic Devices

• 2015 – Analytical Procedures and Methods Validation for Drugs and Biologics


S. P. Piccoli

Meetings and Workshops

• 2013 - Crystal City V AAPS Workshop: Quantitative Bioanalytical Methods Validation and Implementation

• Booth et al., Workshop report: Crystal City V--quantitative bioanalytical method validation and implementation: the 2013 revised FDA guidance. AAPS J. 2015 Mar;17(2):277-88.

• 2015 - PSTC/FDA Scientific Workshop on Assay validation expectations throughout the qualification process

• IMI SAFE-T, FNIH, Critical Path Institute /PSTC

• 2015 - M-CERSI Symposium on Biomarkers in Drug Development

• U Maryland Center of Excellence in Regulatory Science and Innovation, Critical Path Institute, and the FDA


S. P. Piccoli

Meetings and Workshops

• 2015 - Crystal City VI AAPS Workshop: BMV on Biomarkers • Lowes and Ackermann. AAPS and US FDA Crystal City VI workshop on bioanalytical method

validation for biomarkers. Bioanalysis, February 2016, DOI 10.4155/bio.15.251

• Arnold, Booth, King and Ray. Workshop Report: Crystal City VI-Bioanalytical Method Validation for Biomarkers. AAPSJ 2016, doi:10.1208/s12248-016-9946-6.

• 2015 - The Brookings Institute Center for Health Policy – Facilitating biomarker development and qualification: Strategies for prioritization, data-sharing, and stakeholder collaboration

• FDA/CDER, Biomarkers Consortium, CDISC, C-PATH, Hamner Institute, IMI, Merck, NCI, NIH, PhRMA, Tufts University, UNC

• 2016 FNIH Biomarkers Consortium Workshop: Developing an Evidentiary Criteria Framework for Safety Biomarkers Qualification

• FNIH, NIH, FDA, C-PATH, PhRMA, NCI, Industry


S. P. Piccoli

Rationale for PTC White Paper • Characterization and validation of biomarker assay performance is

fundamentally important to biomarker qualification

• Multiple guidance documents have been published for in vitro diagnostic and pharmacokinetic assay development and validation

• These documents contain nearly all of the fundamental concepts necessary for the development and validation of biomarker assays for use in the qualification of Drug Development Tools

• However, the direct application of these concepts has not been codified into guidance for biomarker assay validation


S. P. Piccoli

•

•

•

The Pathway

42

Stakeholder Initial Meeting

Points To Consider White Paper

Public Workshop

Iterative Refinement

White Paper Overview and Terminology S. P. Piccoli

PTC WP Primary Goal

US FDA Scientists

Pharma/Biotech Scientists

The goal of this effort is to define the scientific and regulatory considerations for the validation of soluble biomarker assays used in the qualification of Drug Development Tools.


S. P. Piccoli

Stakeholders

Biomarker Evaluation,

Qualification, Utilization

Industry

Regulatory Agencies

Consortia

Advocacies, Foundations,

Societies

Federal Partners

Academia


S. P. Piccoli

PTC White Paper Consensus Requisites

1. The experimental characterization of the biomarker assays used in qualification (Assay Consideration)

2. The approach to defining the requisite assay performance and

acceptance criteria (Assay Validation Acceptance Criteria)

3. Example of safety biomarkers of nephrotoxicity


S. P. Piccoli

PTC White Paper Key Assumptions

• The validation expectations for biomarker qualification assays are not identical to the expectations outlined for pharmacokinetic (PK) or toxicokinetic (TK) assays in DD.

• The performance characteristics of biomarker qualification assays are in line with the Context Of Use (COU), and ultimately, the clinical application of the biomarker in drug development.

• Although an IVD assay is not required in biomarker qualification efforts, adequate assay performance and validation is essential for support of that qualification’s COU.


S. P. Piccoli


• Biomarker qualification assays are not sufficient for, and are not intended to be used as, de facto substitutes for an in vitro diagnostic device (IVD), i.e., approvals (PMA) or clearances (510(k)) by CDRH.

• Qualified biomarkers and their associated defined assay performance expectations are suitable for use in drug development and regulatory submissions, but are not assumed to be directly acceptable in, or transferrable to, clinical practice, i.e., those laboratories regulated by CLIA.


S. P. Piccoli


• Only singleplex soluble biomarkers measured by Ligand Binding Assays (LBA) and Liquid Chromatography-Mass Spectrometry (LC-MS) technologies are within scope.

• All defined categories of biomarkers are within scope in line with the COU.

• Diagnostic, monitoring, PD/response, predictive, prognostic, safety, susceptibility/risk (BEST Resource https://www.ncbi.nlm.nih.gov/books/NBK326791/ )


S. P. Piccoli

https://www.ncbi.nlm.nih.gov/books/NBK326791/

PTC White Paper Structure

• Overview • Biomarker Qualification and the Context of Use (COU) • Analytical Validation vs. Clinical Validation • Biomarker Assay Validation and Fit-for-Purpose • History of Guidance Documents Relevant to Assay Validation


S. P. Piccoli

PTC White Paper Structure • Assay Design and Technology Selection for Biomarker Assays

• Pre-Analytical Considerations • Analytical Performance Requirements • Assay Performance Key Parameters


S. P. Piccoli

PTC White Paper Structure • Assay Validation Acceptance Criteria

• Accuracy (Relative) or Bias • Analytical Measurement Range (AMR) • Parallelism • Reproducibility • Selectivity • Specificity • Stability (Sample)


S. P. Piccoli

PTC White Paper Structure

• Case Study: Kidney Safety Biomarker Clinical Validation • Panel of six with composite measure output

• Case Study: Liver Safety Biomarker Clinical Validation • Single marker


S. P. Piccoli

to Consider Docum1ent: Scientific and Regulato,ry

Considerations for the Analytical! Vallidation of Assays Used in the

Qualification of Biomarkers in Biologica I Matrices

Biomarker Assay Collaborative Evidentiary Considerations

Writing Group, Critical! Path Institute (C-Path)

Steven P. Picco li, Ne oteric Consulting & Jo hn Michael Sauer, Critical Path Institute

Contributi ng Authors:

Shasli i Am u r, U.S. FDA; Jiri Aubrecht, Pfize r; Ama nd a Ba ker, Criti ca l Pat h Institu te; Robert Becke r,

U.S. FDA; Jen nife r Burkey, Critica l Path Institut e; Ro bert Dean, Eli Lilly; Marth a Donoghue, U.S . FDA; Russe ll Grant, LabCorp ; Steven Gutman, lll umina; Kylie Haskins, U.S. FDA; John Kadavil , U.S . FDA;

Nicho las Ki ng, Crit ica l Path Institute; Jean Lee, BioQualQuan; Vasu m Peiris, U.S. FDA; Afsh in Safavi, BioAgilyj;ix; Shell i Schomake r, Pfi ze r; Meena Subramanyam, Take da; Sue Jane Wang, U.S. FDA


S. P. Piccoli

DDT Qualification Terms Term Definition

Biomarker A defined characteristic that is measured as an indicator of normal biological processes, pathogenic processes, or responses to an exposure or intervention, including therapeutic interventions. Molecular, histologic, radiographic, or physiologic characteristics are types of biomarkers. A biomarker is not an assessment of how an individual feels, functions, or survives.

Context of Use A statement that fully and clearly describes the way the medical product

(COU) development tool is to be used and the medical product development-related

purpose of the use.

Qualification A conclusion, based on a formal regulatory process, that within the stated

context of use (COU), a medical product development tool can be relied upon to

have a specific interpretation and application in medical product development

and regulatory review.


S. P. Piccoli

DDT Qualification Terms Term Definition and Reference Analytical Establishing that the performance characteristics of a test, tool, or instrument

Validation are acceptable in terms of its sensitivity, specificity, accuracy, precision, and

other relevant performance characteristics using a specified technical protocol

(which may include specimen collection, handling and storage procedures). This

is validation of the test, tools, or instrument’s technical performance, but is not

validation of the item’s usefulness;

Clinical Validation Establishing that the test, tool, or instrument acceptably identifies, measures, or

predicts the concept of interest.


S. P. Piccoli

Non-DDT Terms Term Definition and Reference Accuracy Accuracy (of measurement) is the closeness of agreement between a measured quantity

value and a true quantity value of a measurand. (JCGM 200:2012)

Accuracy (Relative) Accuracy is the closeness of the agreement between the result of a measurement and

true value of the measure. In practice, an accepted reference value is substituted for the

true value. Accuracy can also be expressed as %bias. The term "accuracy" involves a

combination of random components (imprecision) and a common systematic error or

bias component (modified from ISO 5725-1);

Allowable Error (EA) The amount of error that can be tolerated without invalidating the medical usefulness of

the result. (CLSI EP21-Ed2)

Analytic Sensitivity Sensitivity is the ability to detect the target analyte within the matrix of interest, and

practically speaking is the limit of quantitation of the calibration/standard curve.

Analytic Specificity Specificity is the ability to assess unequivocally the target analyte in the presence of

components or which might be expected to be present that could have a positive or

negative effect on the assay value.


S. P. Piccoli

Non-DDT Terms Term Definition and Reference Bias Bias is any systematic error that contributes to the difference between the mean of a

large number of test results and an accepted reference value. Thus, it refers to the degree

of trueness between an average of a large series of measurements and the true value of

the measurement.

difference between the expectation of the test results and an accepted reference value (ISO 5725-1)

(of measurements) difference between the expectation of the results of measurement and a true value of the measurand (ISO 17511)

Detection Limit Detection Limit is a measured quantity value, obtained by a given measurement

procedure, for which the probability of falsely claiming the absence of a component in a

material is β, given a probability α of falsely claiming its presence; (JCGM 200:2012);

Limit of Detection LOD is the smallest quantity of an analyte that can be reproducibly and statistically

(LOD) distinguished from the background (including variation in background), or a zero

calibrator in a given assay system. This term is not synonymous with analytical sensitivity.


S. P. Piccoli

Non-DDT Terms Term Definition and Reference Method Development, Lee J, et al. 2006. Fit-for-Purpose Method Development and Validation for Successful

Qualification, Biomarker Measurement. Pharmaceutical Research. DOI: 10.1007/s11095-005-9045-3.

Validation

(Exploratory)

Parallelism Parallelism is the extent to which the dose-response relationship between two materials (i.e., calibrator vs unknown specimens) is constant for the examined range of concentrations. (To confirm that calibrators react in the assay in the same way as the endogenous molecule, CLSI ILA34).

Reportable Range Reportable range means the span of test result values over which the laboratory can establish or verify the accuracy of the instrument or test system measurement response.

(CLI! ’88, Sec; 493;2); Comprised of !nalytical Measurement Range (!MR) (the primary

“linear” range of measurement) and anything done to expand this range (dilution or

concentration).


S. P. Piccoli

Non-DDT Terms Term Definition and Reference Selectivity/Interferenc Selectivity is the ability of the assay to determine the identity of the analyte definitively

e in the presence of the other materials present in the matrix. If the lack of selectivity

comes from a known source, it is referred to as interference; if it comes from an

unknown source, it is referred to as matrix effect (Lee and Hall, J Chrom B 877, 1259-

1271, 2009).

Spike Recovery Spike recovery is the process of comparing the amount of analyte present in a sample

after a standard has been added to and extracted from the sample, as compared to the

true concentration of the standard added. Spike recovery is commonly measured by

measuring the extraction efficiency of the analyte using an internal standard and

showing that it is consistent, precise, and reproducible at more than one concentration.

Trueness Trueness (of measurement) is the closeness of agreement between the average value

obtained from a large series of test results and an accepted reference value (modified

from ISO 17593).


S. P. Piccoli

Terminology Sources • BEST (Biomarkers, EndpointS, and other Tools) Resource [Internet]

Glossary. FDA-NIH Biomarker Working Group, Available at: http://www.ncbi.nlm.nih.gov/books/NBK338448/, Update: December 22, 2016.

• Guidance for Industry and FDA Staff Qualification Process for Drug Development Tools. U.S. Dept. H&HS, FDA and Drug Administration Center for Drug Evaluation and Research (CDER) January 2014. Available at: https://www.fda.gov/downloads/drugs/guidances/ucm230597.pdf

• CLSI Harmonized Terminology Database. Available at: http://htd.clsi.org/.


S. P. Piccoli

http://www.ncbi.nlm.nih.gov/books/NBK338448/https://www.fda.gov/downloads/drugs/guidances/ucm230597.pdfhttp://htd.clsi.org/





61



Session 1a: Assay Design,

Development, Validation &

Pre-Analytical Considerations Meena Subramanyam, Ph.D.

Vice President and Global Program Leader

Takeda

June 14th 2017

62

-


Panelists

• Shashi Amur, U.S. Food and Drug Administration

• Robert Becker, U.S. Food and Drug Adminstration

• Amanda Baker, Critical Path Institute

• Jean Lee, BioQualQuan

63Assay Design, Development, Validation & Pre analytical Considerations

M. Subramanyam

-

Biomarker Qualification Program

• Biomarkers: endogenous entities or molecules.

• The US FDA Biomarker Qualification Program is designed to provide a mechanism for external stakeholders to work with the Center for Drug Evaluation and Research (CDER) to develop biomarkers for use as tools in the drug development process (FDA 2016).

• The goals are to provide a platform to • Qualify biomarkers and make supporting information publicly available. • Facilitate uptake of qualified biomarkers in the regulatory review

process. • Encourage identification of new biomarkers for use in drug

development and regulatory decision-making (Amur et al. 2015).


M. Subramanyam

-

Biomarkers: Assay Design/Development/Validation

• Biomarker Assays typically measure increase/decrease in endogenous level.

• Affected by individual variability in physiology, disease

biology, pathology, co-morbidities, treatment and environmental factors.

• Design/Development/Validation of Assays used for qualification of biomarkers depends on:

• Type of molecules being measured • Context in which the biomarker is being applied in drug

development and in regulatory decision making.


M. Subramanyam

-

What is Covered in the Guidance

• Analytical validation of singleplex ligand, immuno-binding

assays, mass spectrometry, and enzyme based assays.

• Out of scope: analytical validation of immuno histochemistry, flow cytometry, genetics, genomics, imaging biomarkers, and multiplex assays for biomarker qualification.

• General analytical validation principles outlined in this document may also be applicable to exploratory biomarker methods used in clinical development of biopharmaceutics.


M. Subramanyam

-

Biomarker Assay Design/Development/Validation

• Assays for biomarker qualification should be analytically validated to ensure accuracy and reproducibility of data.

• Characterization of assay capability and limitations is critical.

• Measurement errors that could result in biases and affect the biomarker’s predictive accuracy would limit its utility as Drug

Development Tool (DDT).

• Performance characteristics also dependent on COU, application

• COU helps define the fit-for-purpose expectations for assay validation.

• Analytical validation parameters ≠ parameters for PK/TK assays.

• Validated biomarker test ≠ CDRH clearance.

• Assays not IVD and not transferrable as is to regulated clinical practice.


M. Subramanyam

-

Assay Design and Technology Selection • Context of Use - Intended purpose – should be defined a priori.

• Test population (e.g., human (healthy; disease) • Impact of sample acquisition procedures

• Collection timing, methods, transport/storage, sample preservation for stored analyses, contamination concerns (blood in CSF)

• Use environment: • Highly controlled or field-like conditions; High volume or low volume testing • Professionally trained or lay operators; Iterative changes for improved

performance.

• Endogenous levels; prevalence in N/D. • Establishing working criteria is foundational to design and technology selection.


M. Subramanyam

-

Points for Consideration: Assay Design & Development

Assay Features – Characterizing Assay Capability

• Qualitative/SemiQuantitative/Quantitative

• Calibrators/ reference material

• Controls (external, internal)

• Reportable range; Reference interval

• Specimen quantity requirements

• Allowable analytical precision and total error;

• Desired detection sensitivity – upper and lower limits;

putative detection range

• Interference factors in endogenous matrix

• Results Turn-around time

• Batch analysis vs random access

• Singleton Vs. Multiplex assay

• Automation, process software

• Analytical software, user interface

• Waste/hazard containment

• Cost

• Technical support requirements


M. Subramanyam

-

Optimization - Design of Experiment • Checker-board or design of experiment (DOE) approach (fractional

factorial experiments, central composite designs)

• Multi-parameter evaluation: • Minimum required dilution of samples • Assay reagent concentrations • Calibrator levels • Incubation periods, blocking and washing parameters etc., for

optimization of the assay prior to finalizing the assay format. • System Suitability: Regression model for quantitative assays

(polynomial (linear, quadratic); nonlinear models (four or five parameter logistic model, power model)) for the calibration curve.


M. Subramanyam

-

Analytical Performance Requirements: Variation & Bias

• Goal of the study is to determine within-subject (CVI) and between-subject (CVG) variation (Fraser et al., 1997) to define performance

needs.

• Samples from three subjects (between-individual variance, CVG) over 3 days (within-individual variance, CVI) and measuring each

specimen twice (singlicate measure on two separate days to derive

in part analytical variance, CVA.

• Either normal, diseased (two sets of 3 subjects).

• More subjects is naturally optimal if the expected change in the biomarker is small (< 20%).


M. Subramanyam

-

Analytical Performance Requirements:

Total Allowable Error

• Allowable precision (CVA) and bias (BA) are intrinsically related.

• When CVA is large, BA should be minimized and when BA is large, CVA should be minimized.

• Quality specifications for total error (TE) (Westgard et al., 1974; Fraser, 2001 ) computed by addition of bias and precision in a linear manner.

• Bias as an absolute value, no consideration to positive or negative direction of bias.


M. Subramanyam

-

Low Varia11.ce H igh Variance

Analytical Performance Requirements: Variation & Bias

http://scott.fortmann-roe.com/docs/BiasVariance.html


M. Subramanyam

http://scott.fortmann-roe.com/docs/BiasVariance.html

-

Analytical Performance: Desired Goals •Analytical variation (CVA) will add to the “true” test result (Ichihara and Boyd,

2010).

•Optimal: CVA < 0.25*CVI, where CVA comprises ~3% of CVI.

•Desirable: CVA < 0.5*CVI, where CVA comprises ~12% of CVI.

•Minimal: CVA < 0.75*CVI, where CVA comprises ~25% of CVI.

•Analytical bias (BA) is acceptable error associated with a measurement.

•Optimal: BA < 0.125*(CVI2 + CVG2)1/2, would falsely assign a maximum of 3.3% and minimum of 1.8% of subjects outside the group.

2)1/2•Desirable: BA < 0.25*(CVI2 + CVG , would falsely assign a maximum of 4.4% and minimum of 1.4% outside the group.

2)1/2•Minimal: BA < 0.375*(CVI2 + CVG , would falsely assign a maximum of 5.74% & minimum of 1.4% outside the group.


M. Subramanyam

-

Biological Variability

• Serial sample collection from longitudinal study in individual patients within disease context.

• In-study evaluation of total allowable error may be needed to accommodate biological variability.


M. Subramanyam

-

Technology Selection • Selection of technology platform driven by nature of biomarker (protein, nucleic

acid, etc.) and sensitivity requirements.

• Biophysical nature of assay technology and quality of assay reagents will impact absolute and relative measurements.

• Plate- and bead-based assay formats; detection modalities including fluorescence, chemiluminescence, electro chemiluminescence, chromogenic,

mass-spectrometer assessments, acoustic detection systems.

• Various assay parameters compared using fixed set of reagents for biomarker detection with a given technology.

• Normal samples to estimate reproducibility and relative error of back fit concentrations of the biomarker (spiked or endogenous) in relevant buffer

matrix.

• Comparison then extended to disease samples of interest


M. Subramanyam

-

I b'le 2. l.eve:ls ofc rokin derived. from th stim1.1.lated P8MCs1.1pen at ru anm.s_piked inrtopooled. hi.i.man .seru.m toprep re th enmooer1.1:msq11 ·1ity coi'.1Jtro·1 (EQC); lt"o.i~spik~ nd. lov -.. 5pike

m.p' '

'fechnolooy ·1L~2 (p 11l) 1 L,.16 (po.i /ml) IL-17 ----------E QC Hi h-spi ~d Lo,. spik.ed. EQC H. o.i 11-.spiked. Low-.s_pik d EQC H·o.i .s.pik.ed, Low .spiked. ErQC H ig h-.spiked lo, . .s_pi keel. Siroo 2,9!3 732 1416 0..11391 173 035 112 2,,80 0 . .56 174 4.36 0.87 Erenoo. 2,9!3 'NA 'NA 0,1691 'NA 'NA 112 'NA 'NA 'NA 'NA 'NA Mil'li_p X 146,.48 29!2,96 58.,§91 3.46 1692 138 5,60 1120 224 43.i61 8712 17.44 V-_pf 146 . .48 29:2.96 58 .. §91 3.46 i69Q 138 5,60 1120 224 43.61 87.22 17..44 EUSA 'NA 'NA 'NA 173 3.46 0J69 280 5 . .ifO 224 10 001 21.81 4315 'BA'f 146,.48 200.96 ,58._§9' 3.46 t6.00 138 5,60 1120 224 43.161 8712 17..44 Ell 146.48 200.96 73,24 3.46 t6.-92 173 5 .. 60 1110 2,80 43.161 8712 2180 AMMPViBE NA NA NA 2.31 NA NA 11.20 NA NA NA NA NA lmperacer 2.913 732 146, NA NA NA 'NA NA NA NA NA A

'NA - not ap_pltt: 'b snot ev 11.1 ted,

Selection of Technology: Illustrative Sensitivity Assessment

D. Yeung et al. / Journal of Immunological Methods 437 (2016) 53–63


M. Subramanyam

-

As n ·1yrc- I .9ensitivily. in po nmL 1!1.5ing recom.Ifn nt protein spiket:il. in . n 5: ybuffer.

lL-2 lL-15 11..-17 TNFo.

ble 1S Pe.rice m lreq

lL-2 ll-16 lL-17 TNFo.

lm.per c:e:r Mill i_ple.x Erenn Sim

0.46 Ol49 0..05 0.02 Ol73 0.02 0.16 293 0..03 0.02 1.71 0.28

to · for the IL-2 ELISA f. i d 2i rceptanc-ecnler· in sinfi' run _perf'o

.re detection ('FE.AU) in 401 indrvid1..1aJ 1 .pfes.

MiHiplex

98 88 93 1001

Erenn

1001 1001 901

f ill!lrie of libr , rs in FFAD ev I · non .1

Sim ,

95 917 100 100

Arvffi.i1 P ViBE V- plex

0.72 328 0.38 328 2.316

0 .. 62

AMMPViBE V ple.x

0 70 1fiO 53 0

915

EH

1.014 0\42 1 .. 014 1.014

'Ell

0 53 43 1001

EUSA

'NDb 0.32 0 .. 46 0.50

20 100

HAT

0.901 0 .. 12 3 .. 915 0.59

BAT

30 93 5 901

Selection of Technology: Illustrative Sensitivity Assessment

D. Yeung et al. / Journal of Immunological Methods 437 (2016) 53–63


M. Subramanyam

-

Pre-Analytical Factors • Appropriate balance of rigor: fit-for-purpose approach.

• Sample type, collection tube/time/procedures, processing, transportation/handling, storage/retrieval.

• Additional consideration: physiology and/or patient specific characteristics • Age, gender, ethnicity and ongoing diseases. • Factors such as exercise, eating, drinking and medication (Table 3). • Consistency in assay validation, qualification, and post-qualification use. • SOPs, quality control indices, sample acceptance/or exclusion.

• Pre-analytical factors may change across multiple assays and biological matrix .

• Sound scientific expertise and understanding must be utilized for each assay developed for a specific COU.


M. Subramanyam

-

Pia ma v erum B Traditional vs P1·otocol 6001 0 S0001 0 400 0 ~ 0

~ 200 2500 0

0 e ~ 100 0 ~ 300 t:1.2 0 ~' &;- 0

~ ~ -o 0 ,.. 0

a ~ 200 oO !=- ~

0

('j 0 00 0 ~ 0

a.. 0 I 0 rP 0 ~ ~ . . 100 0 • .J.

E 0 0 0 ~ ;e C = -50 0 0 0 ••••••• I:? ••••• • - .............. , ~ .,I ·- 0 't, 0

-100 i'! t, -100 -~ > -~ ~ ~~ ~ ,~ J

' Biomarkers Biomarkers

Pre-Analytics Xiaoyan Zhao, et al., J Immunol Methods. 2012 Apr 30; 378(1-2): 72–80. Published online 2012 Feb 17. doi: 10.1016/j.jim.2012.02.007

Other Publications: Sid E. O'Bryant et al., Guidelines for the standardization of preanalytic variables for blood-based biomarker studies in Alzheimer's disease

research. Alzheimer's & Dementia, Volume 11, Issue 5, 2015, 549–560


M. Subramanyam

https://www.ncbi.nlm.nih.gov/pubmed/?term=Zhao X[Author]&cauthor=true&cauthor_uid=22366959https://dx.doi.org/10.1016/j.jim.2012.02.007

-

Table 1: Approaches for Biomarker Assay Validation

Decision leYel Stage of drug deYelopment Reference Standard

Matrix

Standard and Quality Control Accuracy and Precision criteria

Accuracy** and Precision qualification

Stability eYaluation

DiscoYery/Exploratory Validation

Rank ordering, screening Discovery

Vvhen available, or surrogate

Authentic or surrogate Test parallelism if samples available

Acceptance criteria not needed Established based on evaluation results

Not required

Bench top Scientific judgment

Translational/Partial Validation

Candidate selection Translational Research

Vvhen availab le, or surrogate

Authentic or surrogate matrix Spiked reference standard Consider disease state, multiple donors Test parallelism

Acceptance criteria based on evaluation results and technology-based analytical considerations Native animal/human samples as quality control samples Minimum one run

Collection, room temperature, freeze/thaw, and long term stab i Ii ty as needed Reference standard or matrix stability test with acquired animal/human sam !es

Full Validation*

High risk actionable data Clinical trials

Requires calibrator or reference standard or surro ate Authentic or surrogate matrix Spiked reference calibrator Consider disease state, multiple donors Test arallelism Acceptance criteria based on evaluation results and technology-based analytical considerations Native animal/human samples as quality control samples

Six runs

Collection, room temperature, freeze/thaw, and long term stability Reference standard or matrix stability test with acquired animal/human samples

Staged Validation


M. Subramanyam

-

Assay Design, Development, Validation & Pre-Analytical Considerations Panel Discussion Questions

• What is the difference between biomarker acceptance as part of an individual IND/NDA/BLA submission versus qualification as part of the drug development tools process?

• Does the assay design and analytical validation considerations vary by BEST categories of biomarker (e.g. Predictive, pharmacodynamic/response, safety, prognostic, monitoring, susceptibility/risk or diagnostic biomarker) and/or the context of use (early phase vs late stage clinical trial, used alone for decision making vs used in combination with “gold standard”.

• Are there parameters that need to be evaluated differently? (Are assay validation acceptance criteria different based on level of risk?)

• Can multiple assays/technologies be used to measure the same biomarker as long as they each meet pre-specified minimal performance characteristics – during the qualification process and post-qualification?

• Is there a situation where different technologies/assay design may need to be considered to measure the same biomarker in different disease indication?

• If a biomarker has been qualified for a given context of use, what type of additional assay validation is necessary to expand the context of use, or to qualify the biomarker for a different context of use?

• Is sensitivity more important than selectivity/specificity (or parallelism) for biomarker assays? How should assay design and validation balance positive and negative predictive value? How does this relate to context of use and level of risk?

• Are NPV and PPV different from assay performance measures?


M. Subramanyam

-

Assay Design, Development, Validation & Pre-Analytical Considerations Panel Discussion Questions

• How to consider effect size for biomarker qualification Vs. effect size to measure clinical effect (safety/efficacy) of therapeutic which is typically used for clinical trial design.

• When in the life cycle of drug development or biomarker qualification process should the biomarker assay method be validated?

• Do commercially purchased kits for known biomarkers be subject to the analytical validation process as well?

• Is the analytical validation guidance applicable to biomarker assays that may be run out of various types of testing labs - CLIA labs performing LDTs or IVD tests, research and development labs performing IUO and RUO

tests under GXP regulations?

• Minimal key pre-analytical factors that biomarker scientists must give consideration to?

• In-study validation of biomarker assays?

• Example of using lessons learned from exploratory and pilot in-study validations for assay development and full validation of biomarker?

• Examples of different pre-analytical considerations for different context of use of same biomarker?

• Does the software being used to process biomarker data also need to be validated? (Part 11 compliance?)

• Clinical Benefits and risks of biomarker versus method risk.


M. Subramanyam





84


U.S. FOOD & DRUG ADMINISTRATION Duke MARGOLIS CENTER for Health Policy

BioAgilytix

I CRITICAL PATH \~ INSTITUTE

Session 1b: Assay Performance & Assay

Validation Acceptance Criteria

Afshin Safavi, PhD Founder & Global Chief Scientific Officer

June 14, 2017

85


Panelists

• Steve Gutman, Illumina Inc.

• Robert Becker, U.S. Food and Drug Administration

• Kylie Haskins, U.S. Food and Drug Administration

• Jean Lee, BioQualQuan

86 Assay Validation Acceptance Criteria

Afshin Safavi, PhD BioAgilytix

Challenges of Setting Up Validation Acceptance Criteria

• Determining assay acceptance criteria for biomarker assays is likely the most challenging exercise for a biomarker assay validation.

• Unlike the predefined acceptance criteria established for small and large molecule PK assays, the acceptance criteria for biomarker assays are dependent upon each biomarker’s physiological behavior.

• Even a more difficult question is the nature of the appropriate validation samples.



Goal • The team tried hard to review the past, current publications/meetings and what seems to be considered as

industry best practices to suggest how to set up acceptance criteria for biomarker validation.

“Guidance for Industry, Bioanalytical Method Validation” (FDA 2001)

“Guidance for Industry, Bioanalytical Method Validation” (FDA 2013)

Joint FDA/American Association of Pharmaceutical Scientists (AAPS) Crystal City V, December 3-5, 2013,

(Booth et al., 2015)

AAPS Workshop Crystal City VI: Bioanalytical Methods Validation on Biomarkers, September 2015 in

Baltimore, (Lowes and Ackerman, 2016; Arnold et al., 2016)

“Fit-for-Purpose Method Development and Validation for Successful Biomarker Measurement” (Jean Lee et al. 2006”

Clinical and Laboratory Standards Institute (CLSI) documents: www.clsi.org

• We hope that this section will provoke a good debate such that at the end, we can come up with some specific and some general strategies for setting up biomarker validation acceptance criteria.



http://www.clsi.org/

Angry Sad

Can't you see the difference?

Can the Assay Discriminate Changes?

• As discussed by Lee et al. (2006), the fit-for-purpose status of a biomarker method is deemed acceptable if the assay is capable of discriminating changes that are statistically significant from the intra- and inter-subject variation associated with the biomarker.

• Example, an assay with 40% total error allowable determined during validation may be adequate for statistically detecting a desired treatment effect in a clinical trial for a certain acceptable sample size, but this same assay may not be suitable for a clinical trial involving a different study population that has much greater physiological variability.



File IND

Support formulation development and

clinical studies

Support marketing application

Approval

Monitor product quality and support

~------• postapproval

-Risk/Benefit -Company Practices

changes

Level of Biomarker Bioanalytical Validation

' Letter of ~

•

Qualification \1 Determination \

The Journey of Biomarker Validation from Early Exploratory to Bioanalytical Validation for Clinical Biomarker Qualification

-Risk/Benefit -Company Practices

Biomarker Qualification Process (Key Milestones) - CDER

The assay should be established and analytically validated as soon as possible in the biomarker qualification process. At a minimum, this should be completed prior to submitting the qualification plan to ensure the assay’s acceptability prior to its use in the confirmatory clinical study.



ALIDATED

When is an Assay Considered Validated? Appropriate assay characterization practices must be applied:

Relative Accuracy

Reproducibility

Analytical Measurement Range (LOD, LLOQ, ULOQ)

Parallelism (MRD and Prozone)

Specificity

Selectivity

Stability

The assay must be able to distinguish biomarker changes that are outside of the normal biological variability.

•NOTE: It is desirable to have a well-performing, fully validated assay so that if additional analytical error is introduced into the assay, the biomarker’s performance will not suffer



Accuracy (Relative) • Accuracy is the closeness of agreement between the result of a

measurement and the true value of the measure.

• Accuracy can be expressed as %recovery or %bias, and is also called Trueness or Bias. Established over 6 runs.

• Since frequently “gold” standard material not available for biomarkers, relative accuracy is commonly measured by comparing the measured value of a known specimen to that of a known value of reference material (or spiked sample).

Accuracy = ((Actual value - Measurement) / Actual value) x 100%



Analytical Measurement Range (AMR)

• The ANALYTICAL MEASUREMENT RANGE (AMR) is the range of analyte values that a method can directly measure on the specimen without any dilution, concentration, or other pretreatment not part of the usual assay process.

• AMR validation is the process of confirming that the assay system will correctly recover the concentration or activity of the analyte over the AMR.

• For assays that can test a specimen without dilution (externally calibrated MS assays with isotope dilution):

Analytical Measurement Range = LLOQ up to the (ULOQ * maximum validated dilution)

• For assays which require specimen dilution prior to measurement (Immunoassays using specimen dilution):

Analytical Measurement Range = (LLOQ * Minimum required dilution) up to the (ULOQ * maximum validated dilution)



Analytical Measurement Range (Cont.) K

ey Q

ue

stio

n:

Ho

w d

o y

ou

wan

t to

u

se t

he

ass

ay a

nd

re

po

rt t

he

dat

a?

Qu

alit

ativ

e o

r Q

uan

tita

tive

• LOD: Limit of Detection or analytical sensitivity is determined via

extrapolation of concentrations from a response signal of + 3SD of the mean background signal determined using blank matrix samples (n>10, usually assay diluent).

• LLOQ and ULOQ: the lowest and highest concentrations of analyte that has been demonstrated to be measurable with acceptable levels of total error.

• Data below the LLOQ should be applied with caution.



Analytical Measurement Range (Cont.) • Both the assay and the biomarker’s intrinsic physiological behavior are the primary sources

of variability in demonstrating the utility of a biomarker and its qualification, both of these

sources of error must be taken into account.

• The outlined described here may be used to define a minimal Performance Standard for the biomarker.

Performance Standard is defined by the amount of Total Allowable Error (TAE) for the biomarker at the Decision Level (XC).

PS = TAE at XC

• The biomarker’s minimal Performance Standard can be used as a guide to set criteria for the

acceptability of the Total Error (TE) associated with the assay.



!' :

V : : . . . .

il~i~~~~ Analytical Measurement Range (Cont.) TE is the sum of all systematic bias and variance components that affect a result (i.e., the sum of the absolute value of the Bias (B) and Intermediate Precision (PI) of the biomarker assay). This reflects the closeness of the test results obtained by the biomarker assay to the true value (concentration) of the biomarker.

TE = B + PI

Bias is any systematic error that contributes to the difference between the mean of a large number of test results and an accepted reference value.

Intermediate Precision is the within-laboratory variation based on different days, different analysts, different equipment, etc.

Performance is acceptable when

observed analytical Total Error is

less than the Performance

Standard (TE < PS).

Performance is not acceptable

when observed analytical Total

Error is greater than the

Performance Standard (TE > PS).



Analytical Measurement Range - Take Home Message

Biomarkers with a high degree of biological variability

+ Lower amplitude of response to stimulus

= Require an assay with relatively low Total Error

While higher Total Error would be acceptable for assays with biomarkers that have low biological variability and higher amplitude of response to stimulus.



Remember that parallel lines

never intersect Parallelism • Parallelism is the demonstration that the sample dilution response curve is parallel to

the standard concentration response curve.

• This is necessary due to the fact that in an LBA, a binding interaction is being measured rather than an intrinsic physico-chemical property of the biomarker.

• There is no apparent trend or bias toward increasing or decreasing estimates of analyte concentrations over the range of dilutions when a test sample is serially diluted to produce a set of samples having analyte concentrations that fall within the calibration range of the assay.

• Unlike PK assays, there is no need to wait for incurred samples (study samples) to be available for an initial evaluation of the parallelism of biomarker assays.



Parallel lines Non.Parallel lines

t t Parallelism Evaluation • Tiered approach for Evaluation of Parallelism, Dilutional Linearity, Minimum Required Dilution.

• In the event that dilutional linearity must be performed due to insufficient samples with high endogenous levels, then parallelism studies will need to be performed when samples become available at a later time during the assay development process, and the data added to the bioanalytical validation study as an addendum.

• For this initial assessment, one can screen a series of samples in the proper matrix (disease-state and/or normal) to find several suitable samples, i.e., those with high endogenous concentrations of the biomarker.

• Typically, a minimum of four (though various recommendations from two to ten have been suggested) 1.5 to 2-fold serial dilutions of each sample is performed to cover the entire range of the analytical measurement range.

• Another method is a five-point admixing procedure which would allow for evaluation of sample to calibrator matrix composition evaluation, and would include both imprecision and bias evaluation



pd

Parallelism Data Assessment • As there is not consensus among either industry or regulatory bodies on exact

methodologies for evaluation of parallelism, different schema may be pursued.

• The mean of the dilution-corrected concentration range for all the dilutions that fall within the assay range must have CV less than or equal to the CV set for the biomarker assay.

• For parallelism target acceptance criteria, one solution certainly does not fit all. The industry standard has trended toward having a CV less than or equal to 15% in LC-MS, 20-25% in hybrid LC-MS, or 25-30% in LBA for the mean of the dilution-corrected concentrations for dilutions that fall within the calibration range.

• However, acceptance criteria stringency may be tightened or loosened, as long as the scientific rationale is justified and documented. It is always important to have information on parallelism, but the acceptable degree of parallelism depends on the COU of the assay.



Parallelism Data Assessment (cont.) • Plot observed concentration (response) versus 1/dilution

factor on a log scale and a linear regression is performed

for each sample and calibrator and looking for a slope of 1.

Sample 1 Sample 2 Sample 3 Sample 4

Calibrator

Ave

rage

Bia

s P

er

Dilu

tio

n C

orr

ect

ed

Ob

serv

ed

Dilu

tio

n L

eve

l C

on

cen

trat

ion

C

on

cen

trat

ion

• Plot dilution corrected concentration at each dilution for sample and calibrator versus log scale 1/dilution factor. A linear regression is performed looking for a slope of zero.

• Plot the average bias per dilution level (y axis) versus the dilution-corrected results for each sample (x axis), and test that the slope of the trend line is very close to zero.

1/Dilution Factor

1/Dilution Factor

Sample 1 Sample 2 Sample 3 Sample 4 Calibrator

Dilution-Corrected Concentration



ro C: 00 'iii

~ ....J

Working range

Log dose (concentration)

Minimum Required Dilution and Prozone Effect • The data from the parallelism and/or dilutional linearity studies should be

used to develop the Minimum Required Dilution (MRD) and verified during 140 120method validation. The smallest dilution of spiked matrix that meets 100

acceptance criteria for parallelism and/or dilutional linearity is considered 80 60the MRD. 40 20

0

• Assay should be evaluated for the possible presence of a prozone, i.e., a hook effect, which can produce a falsely low result in a sample that is above the quantifiable range of the assay.

• If possible, the prozone effect should be evaluated in matrix spiked at concentrations that exceed the ULOQ of the assay (up to 100X, if possible).

• Spiked matrix should be diluted to the ULOQ in calibrator diluent, or matrix can be spiked at different levels above and including the ULOQ.

Dilu

tio

n A

dju

ste

d%

Re

cove

ry

Sample 1 Sample 2 Sample 3 Sample 4

0 10 20 30 40 Dilution Factor



to normal variation Upper Control Umlt (UCL)

1~,-----~(Co_ m_m---,----on_ C._u_se~ ) ______ ~

Out-of.•control Point (Specia l Cause)

Time Reproducibility • Method precision and relative accuracy are performance characteristics

that describe the magnitude of random errors (variation) and systematic error (mean bias) associated with repeated measurements.

• Method accuracy, intra batch (within-run) precision, and inter batch (between-run) precision should be established preliminarily during method development and confirmed in pre-study validation.

• However, biomarkers rarely have fully characterized reference standards so these parameters are often established from patient samples or spiked control material in the proper matrix.



Selectivity • Selectivity is the ability of the assay to accurately measure the analyte unequivocally in the presence

of interferences or structurally unrelated components that may be expected to be present in the intended matrix.

• Samples from multiple individuals of normal and target patient populations (such as 10 from each population) should be tested for the endogenous value of the target biomarker in each individual sample. Recovery of the analyte reference standard spiked into each at high and low levels should be determined and calculated by subtraction of the basal value.

• The assay total error may be used as acceptance criteria of spike recovery. A pre-specified sufficient proportion of the test samples should be found to be acceptable.

• Although it may not be needed or only limited experimentations are performed to assess the effect of hemolysis or lipemia on the exploratory biomarker analysis, as the program matures and moves toward full bioanalytical validation for biomarker qualification, the effect of hemolysis, icterus or lipemia on sample analysis should be evaluated, as well as other potential matrix and drug interferents.



Specificity • Specificity is the ability of a measurement procedure to determine only the component

(measured) it purports to measure or the extent to which the assay responds only to all subsets of a specified measure and not to other substances present in the sample.

• For small molecule biomarkers, the exact structure of the target analyte is known, as well as its metabolites and structurally similar moieties in the intended matrices.

• Protein biomarkers may have multiple endogenous forms, with unknown isoforms and/or catabolites. Therefore, specificity evaluation may not be feasible for the large molecule biomarkers using ligand binding methods independently; additional techniques such as mass spectrometry may be required to assess the fundamental specificity of an immunoassay.

• The acceptance criteria for small molecule biomarkers should be similar to acceptance criteria for PK analysis because they follow similar experimental designs.



Stability (Sample)

• Stability under all conditions can be influenced by time, temperature, humidity, the presence of degrading enzymes, the natural half-life of the biomarker, storage conditions, the matrix, and the container system.

• Sample stability is determined by measurement of observed bias to baseline specimens. As time zero (t0) samples are frequently difficult to achieve for biomarkers, one may instead consider the trend of degradation measured at a series of times, e.g. t1, t2, t3.

• Example: 40% degradation might be acceptable for one biomarker but 10% degradation may not meet the need for another biomarker.



Let the Debate Begin



When is an Assay Considered Validated?

Relative Accuracy Reproducibility Analytical Measurement Range (LOD, LLOQ, ULOQ)

Parallelism (MRD and Prozone) Specificity Selectivity Stability



Plot observed concentration (response) versus 1/dilution factor on a log scale and a linear regression is performed for each sample and calibrator and looking for a slope of 1.

• Plot dilution corrected concentration at each dilution for sample and calibrator versus log scale 1 /dilution factor. A linear regression is performed looking for a slope of zero.

• Plot the average bias per dilution level (y axis) versus the dilution-corrected results for each sample (x axis), and test that the slope of the trend line is very close to zero.

1/Dilution Factor

--------- Sample 1 --------- Sa mple2 - ------- Sa mple3 • -------- Sample4

--------- Cal ibrator

1/Dilut ion Factor

Sa mple 1 Sample 2 Sample 3 Sample 4 Cal ibrator

Parallelism Evaluation • Tiered approach

• serial dilution approach

• five-point admixing procedure



Performance Standard is defined by the amount of Total Allowable Error (TAE) for the biomarker at the Decision Level (Xe).

PS= TAE at Xe

TE is the sum of all systematic bias and variance components that affect a result (i.e., the sum of the absolute value of the Bias (B) and Intermediate Precision (P1) of the biomarker assay). This reflects the closeness of the test results obtained by the biomarker assay to the true value (concentration) of the biomarker.

TE= B + P1

Performance is acceptable when observed analytical Total Error is less than the Performance Standard (TE < PS).

Performance is not acceptable when observed analytical Total Error is greater than the Performance Standard (TE> PS).

Setting up Acceptance Criteria



U.S. FOOD

Date post:	12-May-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times