Stability Testing Meets QbD

Oct. 3, 2007
Contributing Editor Emil Ciurczak reports on the recent AAPS-sponsored conference, “Pharmaceutical Stability Testing to Support Global Markets.” Co-sponsored by CHPA, EAS, GPhA and PhRMA, the event was held Sept. 10-12 in Bethesda, Md.

Although this conference was not specifically designed to focus on Process Analytical Technology (PAT) or Quality by Design (QbD), it was the first I have attended where almost every speaker referred to QbD. Of course, drug products are complex organic delivery systems for the API and much of the QbD emphasis is placed on producing fresh batches. It was refreshing to hear that speakers considered the length of time a drug maintained its “deliverability” to be an essential element of its design.

Click here to hear Emil Ciurczak's interview with workshop organizer Saji Thomas (aka associate director for QC operations at Par Pharmaceuticals, Spring Valley, N.Y.). They discuss the industry trends and conditions that necessitated such a workshop, and how Thomas secured buy-in from industry groups including AAPS, EAS, PhRMA, GPhA and CHPA.
The program began with several plenary lectures, then broke up into topical workshops. Mark Staples (MicroCHIPS) began the plenary program by discussing Quality by Design. Now, for anyone who regularly attends PAT/QbD meetings, this was “old hat.” However, for a stability conference, I think it broke new ground. Since the information for both FDA Guidances and ICH documents are on the respective websites, I will not rehash Staples’ talk. I will just comment that I was pleased to see it as the lead paper.

Gary Buehler (Director, Office of Generic Drugs, FDA/CDER) made the expected disclaimer that his opinions did not represent Agency policy, as did all the FDA speakers. The interesting fact covered in his talk was the large number of product recalls based on stability failures. Buehler stressed that QbD should be used to produce a product that delivers the API properly not only when first produced, but for the duration of the life cycle of that lot.

Stability Conditions

Outside the U.S. — Several speakers emphasized that the U.S. is only part (albeit a large part) of the pharmaceuticals-consuming world. When determining conditions for stability testing, “real world” conditions should be taken into account. These include what happens after a bottle of product is opened as well as when stored under less-than-optimal conditions. Indeed, “optimal conditions” may not exist at all in developing countries. Saranjit Singh (NIPER, S.A.S. Nager, Punjab, India) pointed out that India has climates ranging from alpine tundra and glaciers in the north to desert in the west to tropical regions in the southwest. Packaging and stability tests should take all these climatic variations into consideration.

Central America and the Caribbean was the region discussed by Anabelle Castro (Roche Servicios, Heredia, Costa Rica). She outlined the applicable requirements: ICH Q1A (R2) and Q1F, EMEA CPMP/QWP/122/02 (rev 1), ICH Q5C, Executive decree #504 (Panama), Decree #33850 (Costa Rica), RTCA11.01.04:05 (GT, HD, NIC & ES), WHO Working document (QAS/06.179.Rev.1) and others for the selection of batches, etc. Castro covered the large number of guidances and requirements that exist in the region and addressed how they can all be met while still making a profit.

The particulars of stability requirements in the Arab countries were discussed by Abdel Aziz Seleh (WHO, EMRO, Alexandria, Egypt). Seleh shared a number of presentations by researchers from throughout his area. Climatic regions were assigned and accordingly, long-term storage conditions were recommended. Conditions varied somewhat from 25°C to 30°C and from 35% to 65% relative humidity. These variations reflect that the region contains arid mountains and deserts as well as more humid and cooler zones. Thus, the storage will more realistically mirror outside conditions.

In a similar approach, Lucky S. Slamet (NADFC, Indonesia) spoke about the ASRAN zone IV and ASEAN guidelines on stability studies. The major difference from the other papers was that for studies in this region, the upper limit for humidity would be 75% in order to mirror local conditions.

In transit — Outside the normal purview of stability testing would be the time a product spends in-transit/storage at third-party warehouses. Timothy Schofield (Merck & Co., West Point, Pa.) addressed the issue of excursions during transport. Using ICH guidance, Schofield explained several models for predicting shelf life after a temperature excursion. Different regions use different models and methods for prediction. Indeed, he noted that finding the “trigger” for what constitutes an excursion also varies from one location to another.

Schofield described some complex release models and used risk analysis to attempt to recalculate expiry dates. After going over the assumptions and known or assumed storage conditions, he demonstrated some means of recalculation (for example, the Principle of Similar Triangles).

Split tablets and repackaging — Given that I administer split tablets to my dog, this topic caught my attention. Due to HMO costs, a number of physicians prescribe a stronger dose of medicine, instructing the patient to split the tablets, thus saving on prescription costs. Repackaging from larger bottles to smaller containers or even blister packs by distributors may not give the same lifetimes as when products are tested in-house in the original containers.

Vilayat Sayeed (FDA/CDER’s Office of Pharmaceutical Science) gave real world examples of the stability profiles of split tablets. Using Gabapentin and an OTC product with four actives as examples, he showed potency and dissolution results from the studies. As well, he performed thermal analysis tests. His conclusions were that type G1 and G2 tablets were unaffected, while G3 tablets showed a decline in properties. These results were based on sample weight, hardness, potency, moisture, dissolution, and TGA and DSC experiments.


A number of academics weighed in on the subject of instrumentation and, being a “techno-geek,” I found some of the technical papers fascinating. One excellent example was a paper presented by Stephen R. Byrn (Purdue University, co-author Eun Hee Lee). In this paper, he examined the forces that an API might encounter during processing into a dosage form: thermal energy (drying), atmospheric water (RH), physical force (pressure, grinding, compression) and seeding. He went into more detail about the conditions to which an API is subjected, then showed how he predicted stability based on crystal structure analysis.

Byrn showed a number of photographs of caffeine crystals re-hydrating with associated structural changes. He discussed solid state oxidation with examples of changes in crystal morphology due to various conditions (including kinetics of formation). The sheer amount of data presented was amazing. It was clear that Byrn and his colleagues have a handle on the predictive ability of thermal and x-ray analyses. More importantly, Byrn showed how this technology may be used to develop “design space” by including the physical stresses along with other factors when designing a new dosage form.

Designing for Stability

Strongly tied in with instrumentation, this seemed like a novel concept: design for stability, not just for release. Several speakers (including Wei-Qin Tong of Novartis) emphasized choosing the best compromise between solubility and stability when picking a salt form of the drug substance. The old-fashioned approach was to choose the best salt to deliver the drug to the correct site, regardless of how it stayed stabile with time and stress. Under the new paradigm, stressing and assessing numerous salts of an API leads to the best compromise between “good now” and “good later.”

QbD “Design Space” for Analytical Methods

Two papers that caught my attention were about the analytical methodology and its relationship to QbD. The first was by Jianmei Kochling (Vertex Pharmaceuticals, Cambridge, Mass.). Kochling showed some fascinating data on science-based QbD method development. Using a Plackett-Burman Design of Experiments, Kochling outlined the various factors for the HPLC assay (% organic modifier, temperature, flow rate, and percentages of acids added). The results, which took only two man-days, gave a design space wherein the analyst could feel confident that all the analytes would be separated.

The second talk was by Mark Schweitzer (Abbott Laboratories, North Chicago, Ill.). He argued that “fixed in stone” methods for, say, HPLC methods are not in the spirit of QbD. Rather ranges should be described that, when stayed within, led to a quality separation. For instance, instead of 1.0 mL/min, the spec sheet might state “between 0.8 and 1.2 mL/min.” The important fact here is that the “design space” of the analysis has been proven previously. The difference in his idea was how and when these variations could be added to the standard operating procedure (SOP) or New Drug Application (NDA).

Papers of General Interest

Setting specifications for impurity levels was covered by David Jacobson-Kram (FDA/CDER’s Office of New Drugs). He suggested using science- and risk-based approaches in lieu of preset levels (one size fits all). “Find the hazard and define it — is it carcinogenic, teratogenic, neurotoxic, etc.?” he said. “What is the level of exposure: small doses, continual doses, or a single dose? What are the risks for not using the drug?” He recommended using ICH guidelines and science to set practical limits.

Horacio Pappa (USP, Rockville, Md.) discussed the chapter on analytical instrument qualification (<1058>). The most useful part of the talk was the graphic showing where the various specifications within the USP addressed each step of an analysis: <41> Weight and Balance, <31> Volumetric Apparatus, <1058> Analytical Instrument Qualifications, <1225> Validation of Compendial Procedure, etc. He discussed IQ/OQ/PQ and where the USP might be used as a tool for each. It was pretty interesting to hear his perspective, as most QbD conferences do not include USP speakers.

General Comments

Since the “normal” attendees at a stability function have not always been introduced to the common guidances associated with PAT/QbD, I was pleased to see a number of papers addressing these topics. All the ICH guidelines as well as the PAT Initiative and GMPs for the 21st Century were explained.

Even more encouraging was the number of attendees at each session. Although they were being confronted with these “strange, new” concepts, several hundred people attended each talk. As large as I am, I often wished I was big enough to cover more than one talk at a time. Kudos to the organizers and all the speakers; people who know my short attention span know it takes a pretty good conference to hold my interest for three full days.

About the Author

Emil Ciurczak | Contributing Editor