Do you Have Confidence in your Data? Considerations for ... · (OOS) Results in Pharmaceutical QC Waters Pharma & Biopharmaceutical LC-MS Forum . Eschborn, Wednesday June 20. th.

©2018 Waters Corporation 1

Do you Have Confidence in your Data? Considerations for Minimising Out-of-Specification

(OOS) Results in Pharmaceutical QC

Waters Pharma & Biopharmaceutical LC-MS Forum

Eschborn, Wednesday June 20th 2018 Hélène Boiteux


Industry and Regulatory Trends – Background and the need to ensure product quality – What are inspectors looking for? – Lifecycle management of analytical methods

Developing Robust Analytical Methods & Minimising OOS & OOT results

– Key areas where Waters products can help you

Outline


Key Trends in Regulatory Audit (483 Observations)

0%

5%

10%

15%

20%

25%

30%

35%

WrittenProcedures

SpecificationDetermination

Change ControlProcess

Investiagation(OOS)

& documentation

Cleaning Related Calibrationrecords &

Procedures

Stability protocol,statistical data

Method Validation Traininingrequirements

LaboratoryControls

Frequency has grown by 3% in 4 yrs

Laboratory Analytics (EDS 365)

Qualification &

Compliance

Project Mgmt &

Consultation

Nugenesis SDMS

Nugenesis Sample Mgmt

Nugenesis ELN

Training Services

Validation Services

Empower &

MVM

Consumables &

ASR

How can Waters Help?

Instrument Services

Instruments


Insights Gained from US FDA 483 Citations

14.6% 18.3% 18.2%

14.4%

0.6%

0.7% 0.6%

0.4% 3.3%

4.0% 3.2%

3.6%

5.9%

6.9% 4.3%

4.1%

5.0%

3.2%

2.6%

2.7%

2.8%

2.2%

1.9%

1.6%

3.7%

4.4%

4.5%

2.9%

0

50

100

150

200

250

300

2014 2015 2016 2017

Proportion of 483 Observations Related to OOS Investigations

Investigation extent

Failed to reject product not meetingspecifications

No written record

Incomplete written record

Inadequate QC review of records

Incomplete or missing investigation/CAPA

Investigation failures

36%

18.4%

6.8%

22% 22%

18.5%

6.9% 10.1%

10.9%

30%

40% 35%


Global Financial and Regulatory Implications of Poor Quality

Regulatory Overview (from FDA website) Beginning in January 2008, Baxter Healthcare Corporation recalled various lots of heparin, a blood thinning drug, following a spike in reports of adverse events, including deaths, associated with the product. …able to establish a link between a contaminant found in heparin, oversulfated chondroitin sulfate, and the serious adverse events seen in patients given heparin. We have been able to trace the contaminant to 12 different Chinese companies and it has been found in heparin batches shipped to 11 countries.

> 80 recorded deaths in US alone

Recall costs: Regulatory penalty Recall notification Product retrieval Product liability

Baxter stock dropped 16% from Jan 2008 to Dec 2010


Exponential Growth of Recalls Spawns Product Quality Focus

“When manufacturing problems arise in drug manufacturing facilities, drug shortages may follow... 65 % of all drug shortages are caused by manufacturing and quality issues This underscores the need for a safe and reliable drug supply chain.”

http://blogs.fda.gov/fdavoice/index.php/2016/02/modernizing-pharmaceutical-manufacturing-to-improve-drug-quality-ensuring-a-safe-and-adequate-supply-of-drugs/


“I’ll be blunt: In the inspection realm, we find too many, disturbing examples of firms providing us with false information, and doing so deliberately”. “In order for those of us at FDA to do our jobs, we must receive accurate information from firms. Quality metrics based on false data are meaningless”. “I will tell you that the trend for the future has been established today – prevention, risk-based inspections, quality metrics, data integrity, and corporate accountability”.

FDA 2015 – “New Carrots and New Sticks”

http://www.fda.gov/NewsEvents/Speeches/ucm444088.htm

Remarks of Howard Sklamberg Deputy Commissioner for Global Regulatory Operations and Policy U.S. FDA 2015 Annual Conference April 21, 2015 Washington, D.C.


Regulatory Bodies Need to Trust the Data they are Seeing Data integrity guidances

– Focused on chromatography Review of audit trails as in Annex 11 Focused inspections

– Continual training of regulators – Expect to look at the electronic data, not just printouts

Ensuring the bad as well as the good data is available – Specifically for reanalysis and reprocessing

Find the root cause of issues and OOS – Right scaled lab error and full OOS investigations

Data Integrity is Key for Quality Assurance


If analysts are ‘cherry picking’ only the good results If failing results are being deleted, hidden or ignored If samples are being ‘tested into compliance’ Is the instrument functioning properly and providing accurate results? Are the appropriate methods being used?

What are Regulators Looking for?

Culture of Compliance

“Quality means doing it right when no one is looking.” -Henry Ford

Culture of Quality

VS


Perform Testing

Within Spec Report Results

Preliminary Investigation

Error Identified

?

CAPA

Reprocess Batch

Re-analysis

No

No

No

Yes

Yes

Analytical Protocol

Followed?

CAPA (*)

Batch re-analysis

Yes Initiate lab investigation

Phase I: Lab investigation Batch Quarantine

QA/QC to determine the fate

of the batch

Phase II: Full scale investigation

Source of OOS^: Abberation in

measurement Abberation in manufacture

No

Yes

Lab error identified

?

Process investigation

A View of the Regulators Expected Investigation Workflow

(*) Corrective Action Preventive Action ^Out-of-specification

US FDA OOS Guidance link

https://www.fda.gov/downloads/drugs/guidances/ucm070287.pdf


Review of data and records

Demonstrating Ongoing Product Control and Improvement

Analytical method

Supply chain

Manufacturing process

Pharmaceutical Quality System

Analytical method robustness is critical to assessing

manufacturing process or supply chain robustness

- Design space - In-process controls - Identify areas for improvement / change management

Identify/Control variations - Batch - Supplier - Grade

- Analytical design space - Critical factors and interactions - Specifications and “fit” - Lifetime robustness


Adherence to Data Quality Standards Throughout the Product Lifecycle

QC

checks System

suitability tests

Analytical method validation

Analytical instrument qualification

Source: USP <1058> Analytical Instrument Qualification

Evidence that instrument performs suitably for intended purpose

Evidence that analytical procedure is suitable for its intended use

Evidence that the system will perform in accordance with established criteria e.g. UPS<621>

Evidence of ongoing assurance of test suitable performance during test sample analyses

Sensitivity Precision Accuracy Specificity Linearity Range Peak asymmetry


Developing Robust Analytical Methods & Minimising OOS & OOT Results

Laboratory Analytics (EDS 365)

Qualification &

Compliance

Project Mgmt &

Consultation

Nugenesis SDMS

Nugenesis Sample Mgmt

Nugenesis ELN

Training Services

Validation Services

Empower &

MVM

Consumables &

ASR

How can Waters Help?

Instrument Services

Instruments


Analyst Instrument Sample & Method Material Data Review & Calculation

• Procedure details sufficient to produce intended result? • Performance verification –Benchmarking method? • Data review • Training requirement

• Qualification/PM • Leaks, detector performance • Electrical spikes • Instrument setting, parameters & configuration • Missing vial, # of injections • Robustness

• Solution stability (degradation) • Robustness • Retention time shifts • Missing peaks • Unexpected/system peaks • Interference/Distortion (peak shape)

• Contamination from glassware, vials • Reference standard – purity, expiration, proper characterization • Column – robustness, shelf-life, other application use

• Integration parameters • Calculation • Historical data

OOS & OOT Investigation Workflow

Potential Sources of OOS / OOT

With an understanding of root cause, actions can be taken to prevent future events


What is Contributing to Chromatographic Variability?

Instrument

• Pumps/Flow • Connections • Voids • Detector

Sensitivity • Column

Condition

Reagents

• Mobile Phase pH

• Additives • Contamination • Dilutions • Solvent Quality

Workflow

• Mobile Phase Preparation

• Sample Preparation

• Method operation

Inconsistencies in any one category can result in Out-of-Specification (OOS) data Hence the need for benchmarking practices, diligence and well trained staff


Quality instrumentation contributes to repeatable performance Reliable, robust instrumentation = reproducible results Instrument to instrument performance consistency (and across multi-site

operations) Improved resolution facilitates peak integration

1. Ensure Quality and Consistency of Analytical Results


2. Ensure Consistency in Column Packing Materials & Process

Controlling the manufacturing of the column from the very beginning mitigates reproducibility risks

Understanding column packing provides equivalent column performance across all column dimensions

Recommendations! - Minimise the risk of method variation by developing methods on a quality column - Test 3 columns – same chemistry, different batches, using Waters method validation kits to validate long term method reliability

Waters is a primary manufacturer of silica and hybrid particles – from starting materials to finished product

Small 1-2 kg batches are manufactured to closely monitor parameters that impact process variability


Trend in instrumentation and applications – More sensitive method requirements – Smaller volumes and more valuable samples – Customers can’t afford failed assays and re-assay costs

Samples are PRECIOUS and EXPENSIVE – The vial is the LAST surface that the final sample

encounters

3. Don’t Underestimate the Importance of Vials – Autosampler Consumables


4. Be Proactive, Use Quality Control Reference Materials

System Performance Benchmarking – After installation / After maintenance – Reference point – Establish control limits – Compare systems – Record/trend retention factors etc.

System Diagnosis / Troubleshooting – Predictable nature of QCRM – Identifies mismatch between

o Expected – Benchmarking o Observed – Qualification

– Pattern changes in QCRM parameters help target troubleshooting

System Qualification – Routine analysis – Confirmation system qualify for analysis – Comparing results to benchmark data – Monitoring changes in QCRM parameters – Compiling qualification data overtime allows creation of

control chart

Types of Monitoring – System Components

o Pumps, injector, detectors, column Manager, columns – Analyst error – Use QCRM to ensure proper performance when

switching between separation methods


Decision Tree for Choosing Benchmarking Standards

What Separation Technique is being used? HILIC

Preparative Reversed Phase

LC

Analytical Reversed Phase

LC

SFC

HILIC QCRM (186007226) UPC2 QCRM

(186007950)

Separation Method? Separation Method?

Gradient Isocratic

Prep. Chromatography

Mix Standard (186006703)

Autopurification System Standard

(716000765)

Gradient

Isocratic

Neutrals QCRM (186006360)

What type of detection?

UV only & UV/MS MS only

Reversed Phase QCRM (186006363)

LCMS QCRM (ToF) (186006963)

QDa QCRM (186007345)

Quad LCMS QCRM (186007362)


Method development processes can be streamlined to produce the best method in a short timeframe – Screen multiple eluents and columns to identify the most robust method – Benefit from tools that prepare buffers consistently (reduce errors) & check pH

robustness of the method – Use Empower built-in/custom calculation and custom reporting to make the best

decisions wrt method development and avoid manual errors in calculations – Benefit from fit for purpose QbD software and Empower data link – Apply modern technologies to ensure data quality and regulatory confidence

5. Use Fit-for-Purpose Analytical Tools to Develop Robust, Reliable Methods


Streamline Method Development Develop Robust Methods

Applicable to all LC systems • Systematic approach to define

Knowledge Space • Independant from particle size UPLC Benefits - 6x improvement in productivity • Reduce overall development time • Reduce cost per sample • Reduce time per analysis • Improve separation integrity • Improve resolution Compatible with QbD & DoE guidelines (ICH Q8 (R2)) • Using Fusion software to automate the

method development workflow & establish the Design Space


Applying a QbD Approach to Analytical Methods Method Operable Design Region (MODR)

MODR benefits: Flexibility of movement within validated

method operation design region - No need for revalidation within the

design space

Design Space: the multidimensional combination and interaction of input variables (e.g., material attributes) and process parameters that have been demonstrated to provide assurance of quality. Working within the design space is not considered as a change ICH Q8(R2) Guidelines


The Fusion QbD Approach to Method Development

Define Experimental

Design

• Run experiments & process data

• Analyze results & variable effects

Develop Knowledge

Space

• Define best method conditions

• Define method robustness Establish

Design Space

1.

2.

3.

• Select study variables & define ranges – Columns, solvents, temperatures...

• Generate Design of Experiments (DoE) Fusion works seamlessly

with Empower & ACQUITY UPLC H-Class

- Experimental design from Fusion is automatically reconstructed in

Empower

- Sample set and all methods are instantly generated, saving time

and REDUCING MANUAL ERROR

- Processed analytical data is automatically imported into Fusion

for statistical analysis


US FDA - Emphasis on New Technologies

Over the life cycle of a product, new information (e.g., a better understanding of product CQAs or awareness of a new impurity) may warrant the development and validation of a new or alternative analytical method. “New technologies may allow for greater understanding and/or confidence when ensuring product quality. Applicants should periodically evaluate the appropriateness of a product’s analytical methods and consider new or alternative methods.”

VIII. LIFE CYCLE MANAGEMENT OF ANALYTICAL PROCEDURES

https://www.google.ch/url?sa=i&rct=j&q=&esrc=s&source=images&cd=&cad=rja&uact=8&ved=0ahUKEwiO586juePXAhUlmeAKHdMEBycQjRwIBw&url=https://www.fda.gov/AboutFDA/AboutThisWebsite/WebsitePolicies/ucm218116.htm&psig=AOvVaw17VHnnZzMeVkaqSSe5uth_&ust=1512033291545661


Use mass detection to: – Confirm spectral purity in method development – Allow detection of low level process/product-related

impurities (lower LODs than UV) – Ensure detection of non chromophoric peaks – Run LC(/UV) SIM methods for accurate quantification of low

level impurities in routine testing. Better LODs can facilitate integration

Benefit from Mass Information to Help Reduce Investigations and Save Valuable Time

Thereby reducing the risk of problems in routine QC testing


Are your instruments fit for purpose? Is all data coming from qualified & validated systems? Are users well trained (general education and equipment/method-specific)? Are methods validated? Do you have procedural control in place? Where is all of your electronic data?

6. Ensure the Quality of your Data


Traceability in Empower Data & Metadata are Linked

E-Cord: Track info in the

method from batch to batch

The relational db allows full traceability of data – connecting data back to

source... Ensures compliance &

traceability of activities


7. Control Chart Analytical Data to Support Product Lifecycle

QCRM Benchmarking White paper Literature/PN 720004535en Link

o Plot quality and laboratory work performance data

o Confirm when a process, instrument or method is running well

o Warn of a drift towards an out-of-specification result before rejectable material is produced

o Upper and lower control limits (UCL, LCL) can provide notification of attention

o Highlight improved performance after intentional changes

Warning control limit Upper control limit

Typical method performance trending •Retention time range or reproducibility •Peak area range or reproducibility •Peak tailing range •Efficiency •Peak resolution •Peak response •System pressure

http://www.waters.com/webassets/cms/library/docs/720004535en.pdf


System suitability can be automated Create processing method that can flag on limits in the result Custom calculations can be validated and reduce manual or external data processing Configurable user privileges to support data integrity Audit trails give complete view into data and methods to support the quality system,

Empower uses a relational database to connect all actions to one another (e.g. reprocessed and retested data – data integrity risk)

Pre-configured and tailored trend reports can be created to support benchmarking, monitoring and troubleshooting

With a network deployment, appropriate stakeholders can gain timely access to the data to determine root cause and establish corrective and preventive steps to avoid future occurrences.

8. Use Empower CDS to flag OOS Events


Flagging during Review Custom Fields can flag events in the report

Examples of Flagging using Custom Fields


Secure, web-based solution that extends the default dashboards found within Empower

Dashboards to monitor specific aspects of how your instruments are being used, and accelerate investigations by reducing the time and cost of collecting data

Additional customized dashboards are also available, based on your need.

9. Take Advantage of EDS365 Dashboards

EDS365 LETS YOU: • Access critical system usage

information

• Maximize system utilization.

• Manage resources

• Identify training needs

• Identify process improvements.

• Plan for capital expenditures

• Identify opportunities to improve sample turnover time using UPLC Technology

• Identify potential areas of Data Integrity concern


With this report we can you to help to check…. – Are the technical controls sufficient to meet current data integrity regulations? – Is the data protected from unauthorized modification or access? – Are the analysts following company SOPs for multiple processing and manual

integration? – Are the audit trails configured as per regulatory requirements? – Is the data being managed over it’s life cycle in terms of archival and backup?

10. Request a Data Integrity and Compliance Status Report


Proactive rather than reactive maintenance Performance Maintenance is based on knowledge

– A keen understanding of instrument design, component reliability testing, and quality parts manufacturing

– The goal is to optimise instrument reliability by replacing normal wear components before they begin to operate unreliably

Performance Maintenance removes the inherent variability found in Preventive Maintenance that can lead to unbudgeted maintenance costs and lower productivity. – Maximises uptime – Maximises system & laboratory productivity – Improves confidence in the results – Lowers overall cost of ownership by minimizing or even eliminating unscheduled service events

11. Benefit from Waters Performance Maintenance Philosophy


Waters Self-Paced eLearning

12. Reduce Human Error through a Structured Training Program

Control the pace: Learn quickly or take your time

Over 100 modules and 25 tracks to

choose from

Available 24/7: Learn at your own

pace and convenience

Flexible Options Buy modules or tracks to fit your

budget needs

Better Focus: Skip what you know or don’t

need. Repeat difficult course sections

Immediate Savings :

Cut travel costs and eliminate

travel expenses

Affordable: Prices ranging from

$50-$1,200

Certified Developers:

Designed by same classroom based quality trainers

Multi discipline Courses

Empower, MassLynx, UPLC, HPLC, Small and Large Molecules

Fundamental, Essential, & Advanced

Techniques

Demonstration of functionality through hands-on exercises

Knowledge Check Quizzes

Perfect Supplement to Waters

Classroom training


Prevent Failed Results with Waters Quality Instrumentation, Software, Consumables & Support Services

Do you Have Confidence in your Data? Considerations for ... · (OOS) Results in Pharmaceutical QC Waters Pharma & Biopharmaceutical LC-MS Forum . Eschborn, Wednesday June 20. th.

Documents