Outcomes of the Global Bioanalysis Consortium’s Recommendations: Large

Molecule and All Molecules Harmonization teams

Webinar - May 2013

A team

2

A1 – Scope and Regulations

Defined and contrasted Regulations, Guidance and White

papers

Regardless of GxP, recommendations made for performing

Regulated BA with an hierarchical structure – Validations to follow regulated BA principles

– Samples analysis following regulated BA principles, while maintaining

GLP and GCP status for samples

Recommended a draft scope statement for GBC

whitepaper – Clarified scope for regulated bioanalysis

– Included timelines for validations and scope (extent) of validation before

analysis of samples

Created draft glossary and abbreviations for BA

3

Discovery

IND GLP tox

Ph1

Ph 3

Post approval

Ph0

Ph 1 or 2 & Chronic GLP tox

Non-GLP Preclinical or in vitro

Ph 2

Screening, research or qualified methods as desired

Parent drug, prodrug or metabolite in any matrix: qualified method method

Parent drug in plasma: validated method

Parent drug in other matrices: qualified method

Metabolite/Prodrug in any matrix: qualified method

Human vs. animal plasma for MIST: screening method

method

Parent drug & post-MIST* metabolite in plasma: validated method

Parent drug & post-MIST metabolite in other matrices: qualified method

A2 – Tiered Approach for MV

4

A2 – Tiered Approach for MV

Practicality in 483 averse environment

4 tiers proposed (screening, research,

qualified, validated)

When is each tier applied?

5

A3 – Method Transfer and Cross Validation

•Transfer is a process not a cross validation

– Internal: Two P/A batches (4 for LBA)

– External: Full validation & inter-lab comparison (“may” included spiked QC’s)

•Matrix stability not repeated in cross/partial validations

•Cross validation consists of analysis of assessment samples (spiked QCs and incurred samples) assayed using two or more different validated methods

6

A4 – Reference Standards and Reagents

Calibrators and QC samples should be

prepared from a verified stock solution with

proven stability

Altered or surrogate matrix should only be

used if equivalency is demonstrated

between surrogate matrix and authentic

matrix. QC samples should be prepared in

authentic matrix

7

A4 – Reference Standards and Reagents

New reference standard qualification

• Qualified via single partial validation run with L,

M, H validation QC samples (n=6)

Alternate: Stds and QC’s from each lot with

QC’s read from each curve.

8

• Sample collection at testing site

• Storage and shipment from testing site to analytical

lab

• Receipt of samples by receiving lab

• Storage at analytical lab (pre and post analysis)

• Temperature ranges

• Sample management using LIMS

• Sample disposal

A5: Sample Management - Topics discussed

A5: Sample Management - Summary

10

• An audit trail of the location and conditions of the

samples must be maintained from collection to

disposal

• Storage conditions must be monitored at all times :-

Ambient, Fridge, Freezer, Ultracold

(The group has recommendations for the temperature

ranges, furthermore the use of the terms -70 or -80

storage should not be used)

• Temperature excursions outside the acceptance

ranges must be documented together with any impact

• Clinical samples must be disposed of at the end of the

trial

A6 – Stability

Stability in WB not required unless

scientifically called for ; assumes

plasma/serum stability under the same

conditions

Stability in the presence of co-medications

or in unique matrix (e.g. hemolytic) not

required unless scientifically called for.

11

A6 – Stability

Use of freshly prepared calibrators

recommended for long-term stability only

Incurred sample stability not required as a

standard. It should serve to bridge a

possible gap between spiked and incurred

samples, when deemed necessary based

on the physicochemical and /or metabolic

properties of the analyte.

12

A7 – Reanalysis and ISR

Incongruous repeats

• Select prior to PK analysis

• Repeat in duplicate

• Accept if reassay if difference < 30% ;

< 40% for LBA

(Feedback needed, align with ISR %?)

• Report median of 3 results

13

A7 – Reanalysis and ISR

ISR

• Largely a QC effort, scientific return obtained

after first test per unique matrix

Amount of ISR

• 5% of samples for all studies (eliminate tiers)

• Minimum of 20 samples

ISR Failures

• How to deal with individual outliers

14

A7 – Reanalysis and ISR

ISR Selection

• Based on visual inspection, low/high

concentration range of samples, > 3XLLOQ

ISR for Multi-analyte methods

• Select based on primary active entity

• Consideration to other analytes, but not driven

by them unless they represent a major

metabolite or co-med and are largely separated

pharmacokinetically

15

A8 – Documentation

High level table of contents proposed for the

analytical and validation report

Recommendation to reserve a separate

section in CTD on bioanalysis to avoid

confusion and help reviewers locating the

data.

16

Discussion Topics

• Dilution QC’s

• Reinjection reprod./stability

• IS reproducibility

• Integration

• Selectivity/carryover

% LLOQ vs. LLOQ perf.

• System equilibration,

suitability, requalif.

• Tiered MV

• Method transfer/cross val.

• Ref std prep and qualif.

• Stability: ISS, WB, co-meds

17

• Repeat analysis for incongruous

results

• ISR amount, selection and

individual sample failures

Not part of this webinar

18

A9 – Instrument Qualification

A10 – New Frontiers

Not part of this webinar

19

A11 – Biomarkers

Not part of this webinar

20

L teams

21

L1 - LBA Run Acceptance

The team’s recommendations on acceptance criteria employed both in

validation of ligand binding methods and during sample analysis are well

aligned with relevant published regulatory guidance and white papers

Consensus achieved on:

Use of Total Error during pre-study validation to define in-study QC

acceptance criteria

Prepare, qualify and freeze aliquots of standard curve calibrators.

Establish stability by comparing fresh vs. frozen standard curve

calibrators during A&P.

When masking a calibrator is employed, it is strongly

recommended that an objective, step-by-step process be defined a

priori

Quality Controls should be prepared separately from calibrators,

qualified and frozen to mimic the study samples. QCs should

always be used for validation and analysis from the frozen state.

Recommendations:

•Ensure same number of observations in each Precision & Accuracy (P&A) run – number of sets of QCs in each run should be the same

• Aligned with Team L1, which recommends 3 independent sets each of 5 QC levels (LLOQ, LQC, MQC, HQC, ULOQ) in ≥ 6 P&A runs

•Analyst as variable: Number of analysts in validation should be reflective of sample testing practice, noting that:

• Commonly one analyst performs bulk of validation with inclusion of second analyst for a subset of P&A runs

• Possible to justify using only one analyst during validation if study sample analysis will also only use one analyst (small studies)

• If multiple analysts will test samples, recommend at least 2 analysts in validation

L2 - Setting Up Balanced Validation

Design

L2 - Dilution Linearity and Hook Effect

Recommendations:

Approach •Spike samples at ≥ anticipated Cmax; if Cmax is unknown, spike at highest feasible concentration •There should be a minimum of 90% matrix in highest feasible concentration sample •Make dilution series that includes anticipated sample concentrations both above and spanning the curve range

Acceptance Criteria •In-range samples should be within 20% theoretical and precision of the cumulative back-calculated concentrations should be ≤ 20% •If >ULOQ spiked samples read ALQ = no hook effect

Risk based approach

• A case by case risk assessment is needed to determine whether in-study hook effect evaluation should be done

Recommendation:

• The scientist should be mindful of any special characteristics or handling procedures related to the drug and how this might lead to potential hook effect in study samples even when one is not observed during validation

• In study, the scientist should be reviewing the data closely in order to “catch” any hook effect (as well as loss of parallelism) that may manifest

L2 - In-study Hook Effect

L2 - Testing Robustness and Ruggedness Definition • Robustness/ruggedness terms are often used interchangeably and there has

historically been confusion regarding the absolute definition of each term • It is understood, however, that both parameters are indicators of assay

reproducibility under varied conditions • Therefore, any robustness/ruggedness analysis should address the question

of whether the assay will perform well under real life changes in standard laboratory situations

Recommendations • Robustness/ruggedness testing should generally be incorporated into

method development process • Typically includes temperature variations, reagent lots, plate lots, analysts,

instruments, incubation times • Be aware of the needs of your assay and the conditions under which it may be

run, including differences in ambient temperatures in different labs; regional differences in serum/plasma sources, etc.

• Demonstrated during validation by virtue of use of multiple instruments/analysts and typical variations in incubation times

• Cross validation in later stage further demonstrates R&R

L2 - Selectivity

Recommendations

Approach • Test 10 or more individual samples, unspiked and spiked at LLOQ

level (required) and higher level (e.g. HQC) (recommended)

• When possible, for disease indications, selectivity should be performed in disease matrix

Acceptance Criteria • ≥ 80% of unspiked samples should measure <LLOQ

• ≥ 80% (8/10) samples should be within 20% nominal for HQC spikes and 25% for LLOQ spikes

• The same 80% of samples should meet criteria

L2 - Selectivity Lipemic Samples

• Recommendation – risk based approach • The need to perform selectivity assessments with lipemic samples will

be dependent upon the drug, disease indication and assay format

• Typically, performing selectivity assessments with disease matrix samples will address any effects of lipemia which may be present in that population

Hemolyzed Samples – risk based approach

• Recommendation

• The need to perform selectivity assessments with hemolyzed samples will

be dependent upon the characteristics of drug, its target, disease

indication and assay format

• The team actively sought examples where there was an issue caused by

hemolysis to guide when these assessments may be recommended

• Examples gathered to date indicate that issues due to hemolysis are rare and

have not been observed with mAb therapeutics. However, insulin and related

therapeutics may be more likely to be sensitive to hemolysis

L2 - Selectivity – Other Issues

• When endogenous analyte is present

• Choosing standard curve matrix - when possible, screen for low

endogenous level pool

• Employing a subtraction method to evaluate spike recovery may be possible if endogenous levels are measureable (≥ LLOQ)

• May need to sacrifice LLOQ level when endogenous levels are detectable, but not quantifiable (< targeted LLOQ)

• If endogenous level is quite high, then may need to question if your assay range is appropriate

• Routine parallelism assessments currently not being broadly implemented industry-wide (some/few are doing routinely)

• More questions than answers remain around when to perform the assessment, how to perform it, and how to report the data

• Recommendation – risk based approach • The need to perform parallelism assessments will depend upon the

characteristics of the drug, its binding partners and specific assay reagents

• The team sought examples where non parallelism had been observed to guide when assessment may be recommended

• Examples indicate that non-mAb therapeutics, especially peptides, may be more likely to have issues of non parallelism

• No examples yet identified for mAb therapeutics

L2 - Parallelism

L2 - Specificity Testing

Recommendations Approach

• What to test

• Potentially cross-reacting molecule(s); Note: more relevant for non mAb drugs than mAb drugs as selectivity assessments already assess recovery in presence of mg/mL levels of antibodies

• Concomitant medications: As appropriate – test those that are specified in the protocol to be co-administered or stable regimen; No need to test OTC drugs

• Other: Consider potential impact of ADAs, circulating soluble target, etc. on assay performance

• How to test

• Spike maximum concentration anticipated in study into (1) blank matrix, (2) LLOQ QC; (3) HQC

Acceptance Criteria

• Unspiked samples should measure <LLOQ

• Spiked samples should measure within 20% nominal for HQC spikes and 25% for LLOQ spikes

L3: Non-plate based Assay Platforms

Sample testing is run in series (ie Gyrolab, RIA, Biacore) on some platforms

rather than plates

Recommendation:

A run is not limited to ’96’ in this case.

short-term stability data will be used to determine the time required for

‘new runs’ to be started and drift effects if applicable

intermittent QC sets to be added during the sample analysis

A standard curve can be added in at the start of every run.

Carry over should be avoided and the method adjusted accordingly

Singlet analysis of samples can be explored as long as the CV of QC

sets is within an acceptable range developed during assay validation.

As this is the most controversial recommendation, it will require sharing

of experience with validation of larger numbers of QC sets before being

implemented widely.

L3: Assay Platforms

Both transfer between different platforms and multiplexing was also discussed and

surveyed on.

Recommendation:

Method transfer: used of spiked samples to cross-validate platforms was deemed

the most reliable way of method transfer.

Multiplexing best practice: It is recommended that analytes be validated in a

cocktail whenever possible. During sample analysis, if one analyte fails then all

samples should be rerun and mask the previous passing analyte results.

L3: Cell-based assays for PK

More extensive method characterization is required due to the sensitivity of the

assay to numerous factors:

Recommendation for assay characterization:

serum lots, cell passages, plating density, the length of time in culture

conditions to be monitored for assay robustness

sample should be run in triplicate

sample analysis requires that QCs be placed within the plate as well as

along the edge so that any border effects can be determined.

LLOQ, LQC, MQC, HQC, and ULOQ precision and accuracy criteria up to

30% and total error up to 40% may be required.

L3: Cell based Assays for PK

Recommendations for cell line stability:

Cell passage and freezing stability need to be assessed. Within the same

run, QC performance (%RE ≤30%) assessed against calibration standards

in both “fresh” and ‘recently frozen’ cells.

Ranges for parameters such as assay signal, cell growth rate, and viability

should also be considered as appropriate.

New working banks established outside of validation require qualification.

Critical Reagents to be assessed: assay performance with multiple

FBS/FCS lots/sources and different media should be evaluated.

Scope of review was reagents used in PK, ADA and biomarker assays.

Identified four general areas with significant guidance gaps and range of practice:

• Critical reagent documentation

• Critical reagent generated in In-house vs. obtained from a commercial source

• Changing Critical Reagents

• Stability of critical reagents

After extensive discussions, literature and global guidance review draft recommendations

which were first released in June 2012 for feedback from worldwide bioanalytical

community.

Direct comment was solicited through the use of a questionnaire and from conference

presentations, roundtables, responses from EBF, JBF and many individual companies.

The consolidated responses where used to help define the final recommendations and

best practices

Survey results determined only 50% of respondents had documentation for lot changes,

and about 50% having a procedure to address lot changes. Most surveyed indicate

having a procedure for critical reagent production but no procedure for expiry

extension.

L4: Reagents and their stability

L4: Reagents and their stability Changing Critical Reagents

Changing critical reagents represents perhaps the most significant gap and challenge we identified.

Verification is typically done using QCs and criteria used for method acceptance.

The L4 Team recommends testing of new reagents in a functional assay based on a priori assay acceptance criteria. A two-tiered approach for the qualification of these critical reagents is proposed, with the level of testing being driven by the extent of change to that reagent

Examples of minor changes: new reagent lot derived from a previously qualified lot, a new affinity purification of polyclonal sera from the same animals, or a new conjugation using the same protein lot.

Examples of major changes would include antibody lots obtained from new animals, new clones for monoclonal antibody production or new cell lines for the generation of recombinant material

•Recommend five levels of QC samples including the

lower and upper limits of quantitation for PK assays.

For immunogenicity assays the low positive control

close to the cut-point should be included.

•One run for single reagent with minor change.

•Recommend where possible testing in parallel with the

current or original reagent.

•Recommend monitoring instrument response from QC

samples as a tool for gauging assay performance only,

and not as an acceptance criterion.

•Results should be documented. No further validation

is required, if acceptance criteria is met.

•If acceptance criteria are not met, a single failure

might be resolved by a re-preparation of standards and

QCs and reanalysis. However, repeated failures would

indicate the need to prepare a new reagent lot, or new

reagent.

Minor Change - Qualification Major Change - Qualification

The following additional

considerations should be taken

into account for major changes

• Prior to qualification in assay,

the reagent might need to be

characterized for key

properties ( Eg. binding affinity)

•Recommend minimum of

three runs to capture the assay

performance

•If assay acceptance criteria

are not met and/or the assay

performance is altered (Eg.

changed LLOQ), but the assay

still remains fit-for-purpose, a

more extensive qualification or

partial validation may be

required

The regulatory guidance states the expectation on stability shall be assured but

the guidelines do not specify how reagent stability should be assigned and

tested.

Recommendations and best practices:

Assigning test/retest rather that expiry dates to define critical reagent stability

Initial dates can be assigned based on the experimental data, information

from vendor, or experience with similar classes of reagents

Reagents stability testing runs may be included in method development,

method validation, or independent specific reagent stability testing for

intended use

Reagent expiration extension and re-qualification for regulated study support

shall be guided by a defined process which outlines testing and acceptance

criteria

Generation of large amounts of a single lot of a critical reagent reduces risk of

lot changes but requires that large amounts are stored in small aliquots often

at very low temperature to cover the anticipated life of the assay

L4: Reagents and their stability

Reagent stability assignment and testing

Regulations have limited guidance for documentation of assay procedures relating to reagents

and stability. While regulations generally require documentation in the form of an SOP there are

inconsistencies and gaps in what and how to document.

Recommendations:

Documentation can take many forms; can be an SOP or similar document as determined by

your business practices

Identify critical reagents for a specific method, e.g. in the method SOP

Procedure and a priori acceptance criteria for qualification of new reagent lots and location

of this information, e.g. qualification report.

Procedure for determination of reagent stability including;

• How stability will be evaluated with a priori acceptance criteria, frequency, storage

conditions and retesting criteria (failed stability batch).

• How to document stability evaluations, e.g. separate stability batch

• Where to document stability data, e.g. stability report

Procedure to define expiry/retest date, for extending expiry/retest date with a priori

acceptance criteria.

L4: Reagents and their stability

Key Recommendations: Documentation

L5: System Documentation

• Standard Operating Procedure (SOP)

• Installation Qualification (IQ)

• Operational Qualification (OQ)

• Performance Qualification (PQ) (If Significant Software Component)

User Requirements

Validation Plan

Test Worksheets

Validation Summary

Traceability Matrix

Test Incident Log

• Change Control (If Applicable)

• Configuration Only Change Control (If Applicable)

• Configuration Management

• Must be done before assay validation

• Should be included in original PQ for automated system

• No need to re-validate anything already validated

• Includes integration of LIMS in process

• If module added after PQ, there are three options:

o Generate IQ/OQ/PQ documentation as defined

above for automated systems

o Generate a Change Control to another validation

o Perform as part of validation for another system

(including analysis systems)

L5: Validation of a Modular System

L5: Accuracy and Precision Testing

• Scenario 1: Manually validated assay (benchmark

only, out of scope of committee)

• Scenario 2: Assay validated with robotics

• Scenario 3: Manually validated assay along with

assay validation with robotics

• Scenario 4: Manually validated assay followed by

assay validation with robotics LATER

L5 : Automation Practices in Large

Molecule Bioanalysis

System Documentation

SOPs

IQ/OQ/PQ

Change Control

Configuration Management

Validation of a Modular System

Accuracy and Precision Testing

Multiple scenarios exist depending on the assays

L6 - Immunogenicity (effect on PK)

In an upcoming white paper we will provide consensus based

recommendations that are supported by case studies:

• Prior to the development of PK assays the bioanalytical scientist should

consider how the methods may possibly be made ADA tolerant.

• We would recommend the testing of the PK assay for antibody tolerance, but

the evaluation should be done based on potential risk.

• A risk-based approach should be used in conducting investigations, i.e. not

all abnormal PK results require a thorough investigation.

• It is advisable to consider the expected drug therapeutic window during drug

PK method development to gauge the appropriate amount of effort when

examining ADA impact on assay performance.

• PK and ADA data should be considered in combination with other PD and

efficacy markers.

• We advocate covariate PK data analysis that takes into account the effects of

clinically-relevant parameters.