stability testing - who/oms: extranet systems · pdf filestability testing under conditions...

CPH training | May 2016 1 |

Stability testing

Hua YIN


Stability –purpose

Provide evidence of how the quality of an API or FPP varies with time under the influence of a variety of environmental factors such as temperature, humidity and light


Stability –purpose

Understand the molecule

Select packaging

Establish a retest period (API) or shelf-life & storage conditions

Establish shelf life acceptance criteria of Specification

Validate analytical methods


References

Generics Guideline (TRS 970 annex 4, section 3.2.S.7 and 3.2.P.8)

Stability Testing of APIs and FPPs -- WHO Stability Guide in TRS953 (2009) Annex 2:

ICH Q1(A, B, C, D and E)

FDC guideline-- TRS 929 Annex 5 (appendix 3, Table A.1) For PQTm, when there are contradictions, the Generic guideline prevails

Lynda Paleshnuik. Annual Prequalification Assessment Training. Copenhagen, 2010 & 2015


Overview

Stress testing (photostability)

Accelerate, intermediate and long term stability studies

Evaluation of the stability data Summarizing the stability data

Trends, Significant changes, Out of specification results (OoS)

Extrapolation

Examples


Stress Studies

Stability testing under conditions exceeding those used for accelerated testing. Used to:

establish the inherent stability characteristics of the molecule

establish the degradation pathways

identify the likely degradation products

validate the stability indication power of the analytical methods

formulation development, manufacturing process development


Stress testing

When available, it is acceptable to provide relevant data published in the scientific literature to support the identified degradation pathways and degradation products. For example,

If “protect from light” is stated in one of the officially recognized pharmacopoeia, we can assume it is not photostable. Photostability data may not be required if declare “protect from light”, and the container closure system is shown to be light protective.

When no data are available, stress testing should be performed.


Stress testing

No detailed stress testing strategy in Guidelines, except for photostability (ICH Q1B)

experimental conditions and duration may need to be varied depending on the nature of the drug substance

Generic guideline refers to FDC guideline appendix 3, table A.1 (TRS 929 Annex 5) -- as examples


API stability Stress testing – Typical Stress conditions


Stress Studies

How much stressing is enough ? 10-30% loss of API assay (FDC guideline, as well as Generic

guideline) between 5% to 20% degradation (some literature) approximately 10 % degradation is optimal for method validation (e.g.

shelf life assay limit 90.0%-110.0% of label claim)

balance between "purposeful degradation" and irrelevant artifacts (e.g., secondary degradation products).

degradation products not formed during accelerated or long-term stability studies--need not always to be examined


Stress studies: Approach for Assessment: DO

Check if data is provided, either generated by supplier/applicant or from literature references

Check compendial statement, e.g “protect from light”.

Check if conditions are adequate

Check the extent of degradation ≈10% → adequate degradation little or no degradants : verify if conditions are "harsh" enough. If

yes, the API is considered stable.

Conclude on the degradation pathways, stability nature of the API


Stress studies: Approach for Assessment : DON’T

spend excessive time with degradants generated in stress studies if they are not formed in practice.

The impurities/degradants that must be closely investigated are those appearing when stored at long-term and accelerated conditions at greater than (or approaching) the identification threshold (the limit of individual unknowns, 0.1%).


Photostability Studies

Photostability Studies should be conducted on at least one primary batch; see ICHQ1B. Test progressively:

If any stage is photostable, no need to continue, e.g.:

if API shown photostable, then FPP testing is not necessary

API Exposed

FPP

FPP in package


Photostability Studies

Either photostability should be demonstrated or light protection of packaging demonstrated (light transmission data, e.g. USP <661>)

A light sensitive product should be provided in light protective packaging. Be careful to PVC, PVC/PVdC, PVC/PE/PVdC blisters. A opaque carton should be included in the packaging description if necessary with a statement such as, “Store tablets in blisters in the provided carton”.


Overview





Extrapolation

Examples


Stability protocol

Number of batch(es) and batch sizes

Container closure system (s)

Conditions of storage

Testing frequency

Tests and acceptance criteria (reference to test methods)


Selection of Batches (PQTm)

A) a retest period for an API: 3 API batches, at least pilot scale

B) a shelf-life for the FPP:

2 batches of at least pilot scale, OR，

Uncomplicated products* – 1 batch of at least pilot scale + 1 batch which may be smaller (e.g. for solid oral dosage forms, 25 000 or 50 000 tablets or capsules) **

* Uncomplicated FPPs:

e.g. immediate-release solid FPPs (with noted exceptions), non-sterile solutions

** the smaller batch should be representative to the biobatch


Container closure system (s)

Should be the container proposed for storage/distribution including any functional secondary packaging.

Each strength and container type, pack size should be studied unless bracketing/matrixing is applied.

The storage orientation of the product, i.e. upright versus inverted, may need to be included in a protocol when contact of the product with the closure system may be expected to affect the stability of the products contained.

Generally only a consideration for liquids and semi-solids.


Storage Conditions

Climatic Zones Zone I: 21°C/45% RH

Zone II: 25°C/60% RH (subtropical)

Zone III: 30°C/35% RH (hot/ dry)

Zone IVA: 30°C/65% RH (hot/ humid)

Zone IVB: 30°C/75% RH (hot/ very humid)

PQTm:

Long term stability studies should cover all climatic zones if possible, i.e 30°C± 2°C/75% ± 5% RH

Accelerated stability studies: 40°C± 2°C/75% ± 5% RH

* See ICH Q1 for conditions of storage in a refrigerator or a freezer


Testing Frequency

Accelerated : Minimum three points including t0 and tfinal, eg 0, 3, 6.

Long term :0, 3, 6, 12, 18, 24, 36

Intermediate : Four points including t0 and tfinal, eg 0, 6, 9, 12.

For PQTm, when long tem condition is Zone IV, there is no intermediate condition


Minimum data requirements at time of submission

Storage temperature

(ºC)

Relative humidity (%)

Minimum time period

(months) **

Accelerated 40±2 75±5 6

Intermediate * * *

Long-term 30±2 65±5 (API only) or

75±5 (API and FPP)

6

* Where long-term conditions are 30ºC±2ºC/65%±5%RH or 30ºC±2ºC/75%±5%RH, there

is no intermediate condition.

** for 2nd line TB and RH products, 3 months data at time of submission


Stability-indicating quality parameters

Monitor parameters susceptible to change:

Physical: description, moisture;

Chemical: assay, degradants;

Performance tests: e.g. dissolution, disintegration (e.g. dispersible tablets)

Efficacy of additives: Preservative content;

Microbiological testing

Other tests for specific dosage forms, e.g. pH, clarity, particulate matter (for solutions), fineness of dispersion (for dispersible tablets), sterility (for parenterals), reconstitution time for powders


Non Stability-indicating quality parameters

Unlikely to change during storage:

Identity

Content uniformity

Diameter/thickness

Residual solvents

Average weight


What are expected in the stability reports?

General information: Product name, manufacturing sites, date of manufacturing; batch number, batch size, strength, container/clourse (seals, desiccants should be identified)

Stability data in tables or graphs or both

Evaluation – Justify parameters tested (OoS, OoT discussion) – Statistical analysis, if necessary

Conclusions – Proposed retest period or shelf-Life – Proposed storage condition


Overview





Extrapolation

Examples


Evaluation of Stability Data

Review the protocol and report, confirm the number of batches, batch size, packaging, storage conditions, test frequencies, specification.

Review the stability results (details follow)

Notes: Stability of the FPP is product-dependent, not solely due to the API stability. Proceed with caution.



Quality Overall Summary - Product Dossier (QOS-PD) 2.3.S.7 & 2.3.P.8

Use the mock dossier for example

Storage conditions (◦C, % RH)

Strength and batch number

Batch size Container closure system

Completed (and proposed) test

intervals

P. 8 Summary of accelerated and long-term testing parameters (e.g. studies conducted):

http://apps.who.int/prequal/info_general/documents/generic_guide/QOS-PD_Dec2015.docx



Quality Overall Summary - Product Dossier (QOS-PD) 2.3.S.7 & 2.3.P.8

Date of mfg: X95910- Nov 2009, X05172 –Dec 2009, X85110 - Mar 2008

Storage conditions (◦C, % RH)

Strength and batch number

Batch size Container closure system

Completed (and proposed) test intervals

40°C±2°C/75%±5%RH X95910 X05172 X85110 (Biostudy batch)

100,000 tablets 100,000 tablets 150,000 tablets

Blister strip of 2 tabs

Completed – Initial, 1M, 2M, 3M and 6M Proposed – NA

30°C±2°C/75%±5%RH X95910 X05172

100,000 tablets 100,000 tablets

Blister strip of 2 tabs

Completed - Initial, 3M, 6M, 9M, 12M, 18M and 24 M Proposed – 36M and 48M

25°C±2°C/60%±5%RH X85110 (Biostudy batch)

150,000 tablets

Blister strip of 2 tabs Completed - Initial, 12Week, 6M, 9M, 12M, 18M, 24M

P. 8 Summary of accelerated and long-term testing parameters (mock dossier P.8):

http://apps.who.int/prequal/info_general/documents/generic_guide/QOS-PD_Dec2015.docx



Summary of the stability results observed for the above accelerated and long-term studies:

Test Results Description Moisture Impurities Assay Dissolution etc.


Evaluation of Stability Data Summary of the stability results

Example: 40°C±2°C/75%±5%RH, 6 Months. X95910, X05172, X85110

Note: When summarizing, the limits should be included in the stability summary

Test Results

Description (…) Meets specification

Moisture (NMT 0.5%) State range of values or highest value, trend 0.1-0.48%, slight increasing trend. For batch X05172, 0.48% at 6 Month.

Impurities Imp A: NMT 0.5% Imp B: NMT 0.3% Any unkown indi: NMT 0.2% Total imps: NMT 1.5%

State range of values or highest value, trend (for quantitative methods) (Semi-quantitaive TLC-- complies ) Imp A: 0.08 % to 0.32%, increasing trend Imp B: Up to 0.23%, no trend Any unknown indi: up to 0.11% , no trend Total imps: 0.34%- 0.82%, increasing trend

Assay (90.0%-110.0%) State range of values or lowest value, trend 97.8-100.2%, no trend

Dissolution (NLT 75% (Q) in 30 mins) Individual values 81-92%, Average 85-89%, no trend

etc…



When summarizing, the limits should be included in the stability summary

Specifications get revised, sometimes by request. When asking for a tighter limit, the biobatch results and the stability data need to be referenced, and this must include accelerated data results

Always request updated data while assessment is ongoing.

When updated stability is reviewed, it is important to ensure it meets the revised/current limits.


Summarizing impurities/degradants

Impurities (TLC): Results are semi-quantitative, e.g. impurity A < 0.5%, total < 1.0%

Impurities (HPLC): Results must be quantitatively summarized for all specified impurities. This tells: whether there are trends, which impurities are degradants

Focus on the first digit after the decimal point, i.e. 0.X%, variability on the second digit after decimal point is not critical, e.g. 0.02% Vs 0.07% may not indicate an increasing trend due to the variability of method.


Summarizing dissolution data

Dissolution limits should be expressed as a Q value (amount of API expressed as % label content), in line with the harmonized chapters (USP, BP, JP, PhInt)

Q value means three stage testing: Stage 1 (6 units): each unit ≥ Q+5% Stage 2 (6 units): average of 12 units (S1 + S2) ≥ Q, no unit < Q-15% Stage 3 (12 units): average of 24 units (S1 + S2 +S3) ≥ Q, not more

than 2 units < Q-15%, no unit < Q-25% Stage 1 is always tested. Stage 2 is tested only if stage 1 fails. Stage 3 is tested only if stage 2 fails.



Test Results

Dissolution (NLT 75% (Q) in 30 mins) S1: each unit ≥ Q+5% (80%)

S2: Ave ≥ Q (75%) , no unit < Q-15% (60%)

S3: Ave ≥ Q (75%) , NMT 2 units < Q-15% (60%) , no unit < Q-25% (50%)

Results 82-94%, no trend Are the summarized values individual or average values? Does this meet the stage 1 limits?

Example:



Dissolution data - two critical factors to summarize: the individual values (to observe if it meets the limits) the average values (to observe a trend)

The applicant must provide either individual results + average results for each station (e.g. 94, 82, 90, 89, 90, 93, ave. 88)

or, the range of individual results + average results for each station (e.g. individual 82-94, ave. 88)



Test Results

Dissolution (NLT 75% (Q) in 30 mins) S1: each unit ≥ Q+5% (80%)

S2: Ave ≥ Q (75%) , no unit < Q-15% (60%)

S3: Ave ≥ Q (75%) , NMT 2 units < Q-15% (60%) , no unit < Q-25% (50%)

Results 82-94%, no trend

Individual results 80-96%, Ave 82-94%, no trend. √ OR in example where S1 fails: Individual results 78-99%, Ave 82-94%, stage 2

tested and met limits, slight decreasing trend. √

Example:


Overview





Extrapolation

Examples


Trends, Significant changes, OoS

Variation: means you won’t usually have a perfect linear trend (analytical variability, sample uniformity)

Trends: if the majority of stations show a trend (downward, upward), consider it a trend.

OOT: analytical value outside our experience but within the specification (no OOS)

OOS: analytical value outside of the registered specification


“Significant Change”

For an API: "significant change" is failure to meet the specification for any parameter

For an FPP, significant change is any of: Any degradation product exceeding its limit Failure in tests of appearance, physical attributes and functionality test,

e.g. colour, hardness, pH > 5% change in assay from initial, i.e. t0; failure to pass dissolution testing for 12 dosage units (fail S2)


Extrapolation

Extrapolation: Extend the retest period or shelf life beyond the period covered by available long-term stability data

A retest period or shelf life granted on the basis of extrapolation should always be verified by additional long-term stability data as soon as these data become available (commitments)

For long term condition at zone IV (WHO PQTm), extrapolation should be done cautiously. ICH Q1 guidance is for zone II, where there is intermediate data to consider and provide additional assurance.


Extrapolation

ICH Q1E –extrapolation without statistical analysis

No significant change at accelerated condition (6 months )

Little/no change and little/no variability for accelerated and long term data

X= Period covered by long-term data Y= proposed retest period or shelf life

intended for storage at 25ºC or 30ºC

intended for storage at refrigerator 2-8 ºC

Y= 2X but NMT X + 12 months

Y= 1.5X but NMT X + 6 months


Extrapolation

ICH Q1E –extrapolation without statistical analysis

No significant change at accelerated condition (6 months )

change and/or variability for accelerated and/or long term data

X= Period covered by long-term data Y= proposed retest period or shelf life

intended for storage at 25ºC or 30ºC

intended for storage at refrigerator 2-8 ºC

Y= 1.5X but NMT X + 6 months

NMT X + 3 months


Extrapolation

For products for zone IVB, extrapolation should be done cautiously. ICH Q1 guidance is for zone II (25C/60%), where there is intermediate data to consider and provide additional assurance.

For zone IVa/IVb, if accelerated shows significant change, there is no intermediate condition.

Statistic analysis may not available/meaningful when there are limited number of test batches


In case significant changes are observed

Stability is complex and must be considered case-by-case, but various considerations can be weighed to help make a decision, including:

extent of change “Significant changes” that are not failures, e.g. Assay decrease from

100% to 94.0%; Marginal failures, e.g. 89.6% assay for 90.0-110.0% limit; Catastrophic failures, e.g. 70% assay, 20% degradation

time of change, e.g. “Significant changes” observed at 1 months can be considered differently than at 6 months

packaging involved


In case significant changes are observed

Possible responses:

No allowed extrapolation

Shorter shelf-life than covered by long term data.

If long term data is at 30ºC/75%, labeling may be “below 25ºC, avoid excursions above 30ºC

Improve the packaging. If multiple proposed containers and the problem is not observed in all containers, they can withdraw the problem container(s).

A serious problem may require reformulation of the FPP (e.g. known hygroscopic API and wet granulation).



Begin with the accelerated condition, and the intermediate condition if appropriate, then review the long term data

Tabulate and/or plot stability data on all attributes at all storage conditions and evaluate each attribute separately

Identify any trends, significant changes, out of specification results (OoS)

Information of innovator (e.g. packaging of innovator), literature, prequalified products may help in the review

Refer to ICH Q1E appendix A decision Tree for Retest period or shelf life Estimation


Example #1

Problem: Data are provided over the retest period at long-term conditions. Data at accelerated conditions are not available. The applicant states that accelerated data are not necessary because long-term data cover the whole proposed retest period. Is it acceptable?


Example #1

Problem: Data are provided over the retest period at long-term conditions. Data at accelerated conditions are not available. The applicant states that accelerated data are not necessary because long-term data cover the whole proposed retest period. Is it acceptable?

Approach: Accelerated data are not only to support extrapolation, they are also to cover excursions outside the long-term storage conditions. Accelerated data are always required to support a retest or shelf-life period.


Example #2

Data are provided for 12 months at 30◦C/75% and 6 months at 40◦C/75%. No significant change is noted. No/little change and variability. How long could shelf life be granted?


Example #2

Data are provided for 12 months at 30◦C/75% and 6 months at 40◦C/75%. No significant change is noted. No/little change and variability. How long could shelf life be granted?

Answer: A provisional shelf-life of 24 months (12 months + 12 months) can be

assigned. Storage statement “Do not store above 30◦C”. Commitment to continue the statility study should be provided


Example #3

Data are provided for 18 months at 30◦C/75% and 6 months at 40◦C/75%. Significant change is noted at 40◦C/75% at 6 months time point. Data at 30◦C/75% are in compliance with the specification without

significant trends. There is no intermediate storage condition. How long could shelf life be granted?


Example #3

Data are provided for 18 months at 30◦C/75% and 6 months at 40◦C/75%. Significant change is noted at 40◦C/75% at 6 months time point. Data at 30◦C/75% are in compliance with the specification without

significant trends. There is no intermediate storage condition. How long could shelf life be granted?

Answer: A shelf-life of 18 months can be assigned. The shelf-life is based on real time data due to significant change observed at accelerated conditions. Storage statement “Do not store above 30◦C”. Commitment to charge additional batches to stability studies

stability testing - who/oms: extranet systems · pdf filestability testing under conditions...

Documents