Pharmaceutical Stability: Scientific and Regulatory Considerations for Global Drug Development and Commercialization

This article provides a compilation of summaries of many of the presentations at the two- day workshop held in conjunction with the Annual Meeting of American Association of Pharmaceutical Scientists (AAPS) on October 21-22, 2011 in Washington, D.C. The workshop program comprised of twenty presentations encompassing many scientific aspects impacting pharmaceutical stability. The diverse topics included Current FDA thinking on stability practices, phase appropriate method design utilizing QbD concepts, stability indicating spectroscopic methods, design of stability studies for global requirements, strategies for controlling genotoxic impurities in drug products, regulatory perspectives for leachable impurities etc. A session on stability studies of biopharmaceuticals was also included. In other words the workshop topics included most aspects of the “science behind pharmaceutical stability”. The program also included a regulatory round table in which questions were received from the attendees and answered by a panel of regulatory experts from industry.

The presentation summaries have been prepared by the individual speakers, some speakers did not submit the summary of their presentations and therefore those could not be included.

Current FDA Thinking on Stability Practices for New Drug Products (Small Molecules)

Ramesh K. Sood, Ph. D.

The talk was divided into two sections addressing stability considerations at the Investigational New Drug (IND) stage and at the New Drug Application (NDA) stage. He discussed the regulatory and scientific reasons for including the appropriate stability data to support various stages of clinical investigations at the IND stage and to support a NDA for marketing pharmaceutical products in the United States. The CFR [312.23(a)(7)(iv)(a) and 312.23(a)(7)(iv)(b)] states that the IND should provide information sufficient to support stability of the drug substance and drug product to assure their stability during planned clinical studies. It was emphasized that stability data commensurate with the duration of the clinical study are required in all phases of the IND to demonstrate that the drug substance and drug product are within acceptable chemical and physical limits for the planned duration of the proposed clinical studies. He provided some general observations from submitted INDs showing how these submissions addressed these statutory requirements to ensure that the clinical products will remain suitable for their intended use during the course of clinical studies.

The second half of his presentation addressed the need to have appropriate stability data submitted to support an NDA for marketing a pharmaceutical product in United States. It was emphasized that each pharmaceutical product that is to be marketed in the United States needs to have an expiration dating period assigned. This expiration period is assigned to all pharmaceutical products when packaged and stored under the pre-determined circumstances. The data to support such expiration period is generated by studying the stability of the product using carefully designed stability studies. He elaborated on consideration for designing and executing a proper stability study that would generate appropriate stability data to support a proposed expiration period. Finally he presented two case studies which demonstrated how the Agency was able to assign the proposed expiration period in one case and how the data provided in the second case did not support the proposed expiation period.

Design of Stability Indicating Methods Utilizing QbD Concept – a Phase-Appropriate Approach

Dilip R. Choudhury, Ph. D.

The importance of development and manufacture of pharmaceutical products utilizing Quality-by-Design (QbD) approach is well acknowledged by the industry and regulatory agencies. To achieve QbD-based product and process development, it is imperative that analytical methods be designed and developed utilizing a QbD approach as well. The question that is frequently asked is when to start using the QbD approach in drug development and analytical method design since many development candidates do not progress beyond Phase1 or 2 because of failure to meet the clinical objectives.

A unique approach was presented on design of stability-indicating methods utilizing QbD concept in a phase-appropriate manner from the beginning of clinical development. The Analytical Target Profile (ATP) is defined early on to describe the method performance requirements to measure a specific critical quality attribute (CQA) of the drug product. A specific method can be designed to meet the pre-defined performance requirements. Method performance requirements in ATP can be defined in a phase appropriate manner.

The presentation discussed a lifecycle approach to analytical method design and development starting with the early stage method. Critical method parameters are identified during early stage method design and a systematic optimization of critical method parameters is performed utilizing commercially available on-line software. Such optimization has significant advantages over manual optimization of parameters and can be achieved with minimal scientist time. Critical evaluation of data must, however, be performed. Risk assessment of the method is performed at key stages of development. The principle of continuous learning and improvement is applied as the method evolves in parallel with drug development process.

Another advantage of this approach is that since the core method is developed during early development, and the method evolves with the changes in drug product formulation, the final method incorporates the knowledge of the impact of the critical method parameters and formulation and process variables on the method performance. It also provides a convenient means of tracking impurities as a function of formulation and process parameters. The method life cycle parallels the product development life cycle.

In conclusion, applying QbD concept to the design of analytical methods in a phase appropriate manner provides a scientifically sound method at all stages of drug development with good understanding of method risks, critical method parameters and provides better assurance of the robustness and long term performance of the method.

Drug Product Fingerprints: stability-indicating spectroscopic tests

John Bobiak, Ph.D.

The overall quality of drug products relies on management and mitigation (i.e. control) of risks. One approach to managing risk elements considers severity, probability, and detectability of critical events (e.g. form conversion of API, formation of impurities, compositional variations, etc.), and instituting control(s) around each element. Thus, analytical methods fill the need to detect critical quality attributes. Detection, however, is merely one part of the risk management process- a system of actionable controls is responsible for mitigating risk and ensuring quality.

Control strategies for drug products rely on stability data to identify acceptable ranges for ingredients, processing conditions, and storage/ shipping conditions. In one example, degradation product and impurity testing of drug products was waived by proving that: 1) Process-related impurities contained in the drug substance were the only source of impurity content of the drug product, 2) No new impurities or degradants were formed during the manufacture of drug product, and 3) No new impurities or degradants were formed at the long-term storage, accelerated, and stress conditions used in long term stability studies. This testing waiver was complimented with other at-line and online tests to develop a proposal for real-time-release testing (RTRt) of the drug product.

Another example described the use of molecular spectroscopy to monitor form conversion during a drug product stability program. Stressed drug products were analyzed by near infra-red (NIR) as well as powder x-ray diffraction (PXRD), dissolution, and impurity testing. Throughout the investigation, NIR and PXRD identified change of crystallinity at moderate and extreme conditions; the dissolution method did not identify crystallinity changes of the API (BCS I), and impurities were detected in samples of lowest crystallinity. The use of NIR provided detailed understanding of the impact of storage conditions, temperature excursions and packaging types on crystallinity.

In summary, both traditional and emerging techniques offer insight to stability profiles. Stability-indicating tests are an integral part of control strategies for new drug products.

Method Validation at Pre- and Post- Approval Stages Utilizing QbD Approaches

Mark Alasandro, Ph. D.

The use of DOE/QbD method validation approaches was discussed to support stability programs. Such approaches are needed to ensure methods have the accuracy and precision to detect stability changes and provide an understanding of the method variability. Often, method variability alone can suggest stability changes that trigger unnecessary investigations and reformulation activities. Another need is to support pre- and post- approval formulation and process changes without unnecessary method revalidations.

A unique approach was presented using DOE to validate a range of formulations, so formulation changes within this range do not require revalidation. This is coupled with accelerated stability modeling tools to ensure formulation and process changes do not generate new degradation products requiring revalidation. These combined tools minimize the impact of pre- and post-approval changes.

Case studies were presented using DOE/QbD to define a formulation operating range. This can be done without more work than needed using a traditional approach. Other key DOE outputs include a determination of critical method validation parameters that need to be monitored and controlled, such as resolution between critical pairs. A case study was also presented using DOE to assess intermediate precision to ensure there is no increase in method variability.

Another key DOE/QbD output is the Accuracy to Precision model. This shows the balance between accuracy and precision and its influence on product acceptance/failure rates. This can justify moving to new technologies as long as the change meets the accuracy to precision acceptance criteria. An example is discussed starting in early development with a generic gradient HPLC method; and, then going to a product specific gradient method, an isocratic HPLC method; and, finally, to a UPLC or PAT method for product commercialization. This use of DOE/QbD and accelerated stability models provides powerful tools for developing a lean stability program based on sound science and statistical rigor.

ICH Q1B Photostability Guideline – Time for a Revision?

Steven W. Baertschi, Ph. D. et al.

The presentation was based on a recent publication in the Journal of Pharmaceutical Sciences (99:7, 2934-2940, 2010). This oral presentation highlighted some significant and some not so significant deficiencies with the ICH Q1B guideline on photostability, which was published in November 1996 and has been implemented in all three regions (US, EU, and Japan). The presenters noted that since publication in 1996, the guideline has provided a useful basic protocol for testing of new drug substances and associated drug products for manufacturing, storage, and distribution; the authors also noted that the guideline does not cover the photostability of drugs under conditions of patient use. While there were several areas within the guideline that would benefit from revision, a couple of notable issues were highlighted during the talk.

First, the authors noted two significant problems with the Option 2 UVA and visible lamp choices. The cool white fluorescent lamps currently in use in general do not match the spectral power distribution required by the guideline, as specified by ISO10977. Further, the presenters assert that the UVA lamps commonly found in “ICH compliant” photostability chambers often do not meet the requirements of Q1B (where Q1B specifies “A near UV fluorescent lamp having a spectral distribution from 320 nm to 400 nm with a maximum energy emission between 350 nm and 370 nm; a significant proportion of UV should be in both bands of 320 to 360 nm and 360 to 400 nm”). Second, the authors showed that the current recommendation of quinine as an actinometer should be specified only for a specific Option 2 UVA lamp. Further, the timing of absorbance measurements is absolutely critical, since the quinine continues to react after the UVA source has been turned off at approximately one-fifth the rate of when the light is on. The authors provided examples and literature references for their assertions.

The presenters called for the revision process, noting that nearly all of the Quality guidelines have undergone one or more revisions, including the parent Stability guideline, which has undergone two revisions since its first version.

Designing Stability Studies for Global development Programs

Bekki Komas, B.S.

Day one of the Stability workshop included a presentation on Global stability requirements beyond ICH requirements. A recommendation for a global approach to stability and case studies on post approval changes and Emerging Market draft guidelines were shared.

The ICH and WHO stability guidelines were shared at a high level. The discussion included information about countries where actual practice is different than the WHO recommendations, for example Ecuador, Bolivia, Nigeria, Peru and Venezuela.

Recommendations for a global stability approach for drug substance supporting all markets was shared. Additionally, a drug product stability decision tree with consideration of the stability of the product, and the countries where the drug is marketed provided a clear recommendation for long term and accelerated stability conditions. Current emerging market draft guidelines including the draft ASEAN Stability and Variations guidelines and challenges with those guidelines were outlined for group discussion.

A case study for CMC Post Approval change with technology transfer from development to a commercial site and included two multivariate analysis models for stability and design space. Acceptance of the risk based approach and prior knowledge resulted in stability requirement waivers and enabled a different approach to process validation. Key messages from the presentation included that regional stability guidance exists for emerging markets and may be different than what is published in the WHO Stability Guideline. Science and risk based approaches are not always accepted. Some new guidelines indicate an acceptance of stability commitments instead of upfront data for process changes. The ICH Global Cooperation Group which includes 6 regional harmonization initiatives APEC, ASEAN, Global Cooperation Council, PANDRH, SADC East Africa Community are making improvements towards harmonization.

Artifactual Formylation of the Secondary Amine of Duloxetine Hydrochloride by Acetonitrile in the Presence of Titanium Dioxide: Implications for HPLC Method Development

M. J. Skibic, Ph. D, et al.

This presentation was based on a recent publication in the Journal of Pharmaceutical and Biomedical Analysis (53, 432-439, 2010). The authors showed that duloxetine hydrochloride, a secondary amine containing pharmaceutical, currently marketed as Cymbalta, undergoes N-formylation as an artifact of sample preparation prior to HPLC analysis for impurities. They showed that the reaction is catalyzed by sonication and/or light in the presence of titanium dioxide and is proposed to occur via a hydroxyl radical-initiated mechanism. The proposed mechanism was supported by controlled sample preparation studies with deuterium-labeled acetonitrile, and LC/MS studies showed incorporation of one deuterium into N-formyl duloxetine, proving that the carbon of the formyl group was from the methyl group of acetonitrile. This artifactual reaction can be eliminated or minimized by the replacement of acetonitrile with methanol, or by simply adding at least 10% methanol to the sample diluents. The authors provided rationale for the use of methanol using computational chemistry to show how methanol is able to sufficiently scavenge the hydroxyl radicals to prevent the oxidation of acetonitrile. The authors discussed how this discovery is broadly relevant since sonication is commonly used to aid dissolution of pharmaceuticals in acetonitrile for HPLC analysis, titanium dioxide is a commonly used excipient, the amount of light found in modern analytical laboratories is sufficient to cause the reaction to occur, and secondary amines are present in the structures of many pharmaceuticals.

Strategies for Controlling Genotoxic Degradants in Drug Products: Mitigation Strategy and Case Study

Brian W. Pack,Ph. D. et al.

This presentation focused on genotoxic impurities (GTIs) in drug products, a topic currently receiving significant attention in the pharmaceutical industry. The EMA Guideline on the Limits of Genotoxic Impurities (2006) recommended that potential GTIs should be identified and when there is a structural alert, a bacterial reverse mutation assay should be conducted. In addition, it states potential degradation products of the drug substance and drug product should be considered for genotoxic potential with little guidance to determine those degradation products that may be reasonably likely to form. The authors outlined a strategy implemented at Eli Lilly that hinged upon well-designed stress studies to understand the most probable degradation pathways that a compound may undergo, thus limiting the number of potential degradation products to assess for genotoxic potential. Internal Lilly data was presented that illustrated that degradation products observed on long-term stability were a perfect subset of the major degradation products observed during stress degradation studies (n=15 drugs).

If the potential degradant had an alerting structure and was Ames positive, they recommended the development of an analytical method with an LOQ at 10% of the threshold of toxicological concern (TTC). They proposed that if the degradation pathway was inactive (more than 10% TTC) in the drug product on long-term stability that appropriate due diligence had been demonstrated and there is no risk to patient safety. The proposed strategy is novel in that the assessment of potential degradation products (i.e., those derived from stress degradation studies) is dictated not by subjective judgment of what degradation pathways might likely be active, but rather by well-designed stress studies. A final point made by the authors and supported by a case study was to leverage the information that can be gleaned from using Arrhenius predictions from accelerated stability studies in order to quickly screen multiple formulations in order to predict the GTI levels at the end of the recommended shelf life. They advocated that this approach enables formulation and analytical development to make informed decisions around formulation, package, and storage conditions as they relate to GTI formation.

Dissolution failure and failure mode analysis of a Hot melt extrusion product – a case study

Saji Thomas, M. Sc.

The regulatory guidance and the CFR requirements were discussed as an introduction to the main theme of the talk. The high lights of the Barr decision along with FDA’s Guidance on investigating OOS results for pharmaceutical products were discussed. Determining root cause and establishing CAPA is the key to successful OOS investigation. Failure mode analysis tool like Fish bone diagram was discussed as a mean to identify the root cause. The step by step investigation of a dissolution failure of a melt extrusion product was discussed as case study. The product failed at 6 months at accelerated stability condition. Since the initial investigation did not result in identifying the root cause it was decided to evaluate the robustness of the method. A five factorial/eight experiment design was used to evaluate the robustness of the method. The method was found to be very sensitive to the salt added in the dissolution media. A normal probability plot was used to assess robustness of the method using the data generated from the DOE.

Other analytical techniques had to be used since the evaluation of the method did not conclusively prove the root cause of the failure. The product was analyzed using Solid state NMR, Environmental electron scanning microscope, time of flight-secondary ion mass spectroscopy (TOF-SIMS). Products cured at 50c were used for the analysis. SSNMR analysis showed that stearyl alcohol in the formulation if behaving differently after curing the product. A white coating on the product was observed when the product was viewed in an ESEM. The white coating was identified to be stearyl alcohol by TOF-SIMS.

Based on a scientific investigation it was concluded that the dissolution failure was due to the stearyl alcohol migrating to the surface of the product when exposed to heat.

Stability Testing to Support Distribution of Temperature Sensitive Pharmaceuticals

Robert H. Seevers, Ph.D.

All drugs are, to some extent, temperature sensitive. Stability testing makes it possible to determine the appropriate long term storage condition and to evaluate the risk to product quality posed by exposure to temperatures outside that condition. The ICH stability guidances primarily focused on storage conditions and shelf lives for drug substances and drug products. This presentation discussed the other types of testing which may be done in addition to the standard ICH conditions that permit a stability budget to be created for a pharmaceutical.

A stability budget is created using data from long term and accelerated testing such as that provided for in the ICH guidances, but adds freeze-thaw and temperature cycling studies to create a fuller picture of a drug’s susceptibility to changes in temperature that it may experience during the entire distribution process from manufacture to transport, to storage, to end use. The concept of time out of refrigeration has been used to allow for room temperature operations such as packaging and labeling on refrigerated products. This presentation showed how this concept can be expanded to the entire distribution process.

Effective Approaches for Conducting Method Transfer

Kim Huynh-Ba, M.Sc.

Analytical procedures are critical to determine quality of pharmaceutical products at release and throughout their shelf life. Kim Huynh-Ba discussed several options to transfer analytical procedure from one laboratory to another laboratory and effective approaches to be considered when conducting method transfer. Method transfer is a process of which an analytical laboratory is qualified to perform a testing procedure, as method transfer activity is required as part of method validation. This training of performing an analytical procedure is done by The Transferring Lab to The Receiving Lab. It is well documented that there are four options that can be used for this purpose. Three of these four options involve lab activities and the selection depends on the validation status of the analytical procedure and/or availability of lab personnel. These are: Comparative Testing, Co-Validation and Re-Validation. The fourth option is to determine if the Transfer Waiver can be done. Several references including USP new General Chapter <232> 1224 — have suggested this option. Based on the knowledge, ability and experience of the Receiving Site, additional testing may not be needed to qualify The Receiving Lab to perform certain analytical procedure.

As all other cGMP activities, transfer activities are conducted according to a protocol with predetermined acceptance criteria. This protocol should be agreed upon by all sites involved. Elements of a transfer process were also discussed. Depending on the types of labs involved, certain SOPs such as data review, data reporting, OOS or OOT investigation from both sites should be discussed prior to the transfer. Conclusion of the transfer must be documented with all data reported including any representative chromatogram or spectra. It is important that validation data are available to the Receiving Lab as part of the Transferring Background Package and The Receiving Lab must conduct a thorough gap analysis before any testing can be done. It is also advised to avoid cGMP materials to be used for transfer activities to avoid issue that may arise with data generated.

Roundtable Discussion – Interesting and Challenging Stability-Related Regulatory Questions

Panelists: Stephen Colgan, Ph. D.; Robert J. Timpano, B. S.; Ganapathy Mohan, Ph. D.; David Lin, Ph. D.

Day one of the workshop concluded with a Roundtable discussion that focused on Regulatory Queries. A quick show of hands at the beginning of the Roundtable indicated that the areas of most interest included Science and Risk-Based Approaches to Stability/Shelf-life and Stability Protocols.

It was noted that the FDA has been advocating biorelevant specifications, but do not appear to embrace biorelevance when it comes to stability protocols. An example is the mandate to include water content in post-approval proposals when the science clearly indicates that water content has no relevance to stability or biorelevance. The panel noted that there is no guidance in this area, but encouraged the audience to use science to support stability testing protocols. It also was noted that scientific engagement with the regulators in the ICH regions would be welcomed, this is not the case for most of the Emerging Market regions.

The discussion continued with a discussion of science-based stability protocols that should only focus on the relevant product attributes. This approach would be leaner, and would protect the patient as well as non-lean protocols. The panel noted that the Quality Target Product Profile (QTPP) could be leveraged to advocated Lean Stability Protocols.

The most enthusiastic discussion of the Roundtable focused on a liquid product in a semi permeable or non permeable container and whether a matrixed stability protocol would be suitable for containers stored upright, upside down, and lying down. The panel noted that developing a matrixed protocol can be difficult and recommended that a statistician be engaged. It also was suggested to review this type of protocol with the Regulatory Authorities before execution. From a scientific point of view, the panel noted that a Stability Risk Assessment would help define what attributes should be monitored on stability and whether a matrixed approach should be considered. For the question on container orientation, the stability scientist should determine what attributes may be affected by the container orientation. If there is a medium or high risk that the orientation would have an affect on a potential shelf-life limiting attribute, a matrixed approach would not be advisable. In the end, everyone agreed that science and a knowledge of the regulations should be the drivers when stability protocols are being developed.