Continuous Process Verification for a Continuous Process Verification for a L P d t F Sit T f L P d t F Sit T f Legacy Product; From Site Transfer Legacy Product; From Site Transfer Through Post Marketing ChangesThrough Post Marketing Changes

Richard SchichoRichard SchichoJanuary 28, 2013

AgendaAgenda

Process Development/Manufacturing History

Site Transfer to “Paperless” Plant

Tech Transfer and Process Performance Qualification

Continuous Process Verification and Qualification of Post Marketing Changesg g

Process Development/Manufacturing HistoryProcess Development/Manufacturing Historyp g yp g yCommercial process developed at launch site > 10 years ago with limited DOE process> 10 years ago with limited DOE process characterizationProcess transferred to two CMO’s post-approvalProcess transferred to two CMO s post approval to support commercial demandTech Transfers included facility-fit and other minor changes, supported with analytical, non-clinical and clinical PK comparability studiesT h T f t l l t i l d dTech Transfer to paperless plant included facility-fit changes, supported with analytical comparability and enhanced process data package

Tech Transfer to Paperless PlantTech Transfer to Paperless PlantppGreenfield plant designed with highly automated and integrated electronic systemsand integrated electronic systemsElectronic batch records and QC laboratory methods integrated with business systemsg yBatch context database system with > 4500 points configuredSix production bioreactors with a single downstream trainIQ/OQ/PQ studies, kla studies and non-GMP engineering runsC bi d d t k t t i l fCombined data package to support equivalence of bioreactors and N=3 PPQ study

Paperless Plant Components

ERP

rtal

genc

e P

or

Mfg EngineBatch Query/

ReportsKPI Dashboards

LIMS

Methods Execution

urin

g In

telli Mfg Engine Reports

eBatch RcdEnterprise Data Historian

Enterprise Reporting

Maintenance & Calibration

Doc. Mgmt

Man

ufac

tu Batch Engine(S88 Recipe)

Common Data

Training Records

Advanced Analytics

Ent

erpr

ise

Plant Floor (DCS/PLC, BAS, PAT

System/Devices)

DataHistorian QMS/CAPA

eChange Control

E

Plant Execution and Transactional

Mfg Operations Analysis & Reporting

Supporting Business Systems

Interface Interface -- MES/ERPMES/ERP

Interface Interface –– MES/LIMS/Lab NotebookMES/LIMS/Lab NotebookMES LIMS Lab Notebook

Start Recipe Execution

Generate Sample ID’s

Start

Initiate LIMS Job

R t S l ID

Collect Sample

Processing

Record LIMS Sample and

Send ID

Request Sample ID

Return Sample ID

Start Assay Preparation

Send SampleReceive Test

RequestTest Request

Conduct Assay & Approve

Results

Continue Processing

Return Results if Required

If Material

Send Test ResultsResults

Update Genealogy

Consumption Send

GenealogyProcess Order

Complete PFD for Process Unit

Maint Check

ProcessMaint Check Maint

CheckMaint Check

Room Check

Condition Checks Material Checks

Check

HygienicPreparation

HygienicPreparation

HygienicPreparation

HygienicPreparation

Process

Preparation

Process

Process HygienicPost

Process

H i i

Process

HygienicProcess

Process

HygienicPost

HygienicPost

HygienicPost

HygienicPost

HygienicPost

Post

Identifying Resuable ‘Patterns’

Process Performance QualificationProcess Performance QualificationQQThree PPQ runs operated in a single production bioreactorbioreactorProcessing from vial thaw through to drug substance fill under protocol, including media gand buffer preparationDesignated Validation Parameters throughout ththe processExtended characterization testing of drug substance including side by side testing ofsubstance, including side-by-side testing of comparator lots from approved siteLong term, intermediate and acceleratedLong term, intermediate and accelerated stability studies

Continuous Process Verification and Continuous Process Verification and Qualification of PostQualification of Post--marketing Changesmarketing ChangesQQ g gg gQuality Systems

Complaints and Adverse Event ReportingOOS, Market life stability studies, process d i ti (i l di i ld i ti )deviations (including process yield variation)Facility/equipment maintenance and calibration programscalibration programs

Statistical Process Control ProgramN l ti it iNear real-time process monitoringAbility to perform ad hoc queries to investigate statistical trends shifts or outliersstatistical trends shifts or outliers

Statistical Process Control ProgramStatistical Process Control ProgramggEncompasses in-process, release test data and release test method performance parametersp pDevelopment of >400 control charts, PPs, CPPs, CQAs, system suitability and KPIs

Temporary limits set after 10 lots and permanent limits set after 30 lots; Process Capability

A d d i f l h i h ilAutomated updating of control charts with email notification of statistical alertsInvestigation of outliers shifts and trends toInvestigation of outliers, shifts and trends to identify special cause variationCompilation of data for annual reports andCompilation of data for annual reports and assessment of validated state of the process

Statistical Process Control ProgramStatistical Process Control ProgramggCross-funtional team including Mfg Technology, Mfg QC QE Validation and AutomationMfg, QC, QE, Validation and AutomationWeekly meetings for review of data trending; investigation of statistical alert eventsinvestigation of statistical alert eventsDevelopment of CAPA and continuous improvement opportunities

Potential changes to reduce process variability are evaluated via Change Control

t di th d f dditi lsystem regarding the need for additional Process Characterization and/or Qualification

Process HierarchyProcess HierarchyProcess HierarchyProcess Hierarchy

Chromatography NodeChromatography Node

Bioreactor NodeBioreactor Node

Data Model Capabilities Data Model Capabilities –– Control Charts Control Charts

Control Charting of all Critical Process Parameters with automated e-mail alerts for excursionsOver 400 charts refresh on a daily basis and generate alerts via e-mail including Upstream, Downstream and Drug Substance data

Data Model Capabilities Leveraged Data Model Capabilities Leveraged –– Transition Transition AnalysisAnalysis

Data Model Capabilities Leveraged Data Model Capabilities Leveraged –– Raw Materials Raw Materials AnalysisAnalysis

Due to small groupings of batches that utilize the same raw material lot Segmented chart shows raw material high and low performance within one lot of the raw material This demonstrates that this raw

Segmented chart shows also highlighted raw materials that had a potential impact on the bioreactor performanceg p g

numbers, segmented charts were used in the analysis.

The raw materials analysis eliminated many of the raw materials that were not

within one lot of the raw material. This demonstrates that this raw material is not the main contributing factor for the performance.

potential impact on the bioreactor performance.

y ythe primary factor effecting the production bioreactor performance.

In the initial stages of the investigation, some raw materials were identified as g g ,a potential root cause of the performance shift. Changes in the planned raw materials for future batches were made to mitigate the potential impact of the identified raw materials.

Qualification of PostQualification of Post--approval Changesapproval ChangesQQ pp gpp gReview and approval of proposed changeIdentification of key parameters related to potential impact of process changeP d ti f “t t” ith hProduction of “test” runs with process changeStatistical analysis of impact of process change including assessment of Process control chartsincluding assessment of Process control charts

Qualification of PostQualification of Post--approval Changesapproval ChangesQQ pp gpp g

ConclusionsConclusionsHighly integrated electronic systems

Ensure control of process and analytical method executionEnable the collection of data in real timeEnable the use of sophisticated analytical tools f ti it ifor continuous process monitoring, process change qualification and periodic assessment of the validated state of the process

AcknowledgementsAcknowledgementsgg

Joe ShiminskyPete Mikkonen