Send to printer »

Feature Articles : Oct 15, 2007 (Vol. 27, No. 18)

Biologics Force Changes in Stability Testing

Multiple Approaches Are Necessary to Ensure Complete Information
  • Gail Dutton

The introduction of biologicals into our drug armamentarium is complicating stability testing. For these complex molecules neither physicochemical data nor bioassays alone provide the complete picture, so both are usually required.

“You can’t tell everything about molecules by normal physicochemical techniques, especially higher-order structures,” notes C. Jane Robinson, Ph.D., principal scientist, division of therapeutics, National Institute for Biological Standards and Control (www.nibsc.ac.uk), and one of the speakers at Informa’s November meeting on stability testing in Prague.

“Different assays assess the integrity of different properties of the molecule,” she says. You need a bioassay to measure biological activity. For example, degradation products, though usually less active, may still have biological activity and sometimes may be enhanced. “Physicochemical assays may also provide complementary information,” Dr. Robinson adds.

“Bioassays are complex systems and are subject to a lot of variables you can’t necessarily control and sometimes aren’t even aware of, thus affecting the response,” notes Dr. Robinson. Therefore a preparation must be compared to a reference standard of a similar material so changes will affect both materials equally.

Unless the test product and the reference standard are functionally similar, “it isn’t possible to calculate a meaningful relative potency.” Although they need not be identical, greater similarity decreases the chances that changes in procedure will affect them differently.

“Ideally, you want to work on the same standard through the whole drug development and marketing lifetime, though this often isn’t possible,” Dr. Robinson points out. “If a standard must be replaced, it should be planned in time to allow adequate calibration against the previous standard.”

Testing Reference Standards

Ensuring the homogeneity and stability of the reference standard is also important. Bioassays often require a smaller quantity of material than do most physicochemical techniques, so the bioassay reference standard is commonly prepared in smaller aliquots, which may affect stability.

“Moreover, if a subsequent batch of vial or ampoules is prepared, it will have a different handling history,” Dr. Robinson says. If the batch of standard is split between storage locations, periodic testing is needed to show homogeneity.

Many methods are available to stabilize biologicals, so the right combination must be identified for each product. However, the in-house bioassay reference standard must then also be monitored for biological activity. “If the standard is changing, how can you tell whether the activity of your product is changing?” she asks.

For some products, external reference standards exist for biological activity. These can be used to calibrate in-house standards and monitor their stability. “However, for some materials, you won’t have an international reference standard,” Dr. Robinson notes. In those cases, accelerated degradation studies may be used.

The activity of samples stored at high temperatures predicts stability at lower storage temperatures, but results must be interpreted cautiously, she adds, because the degradation pathways at high temperatures may not reflect those at lower temperatures. Specifically, “degradation rates calculated from high storage temperatures of short storage times tend to underestimate the stability.”

Parallel stability tests offer another possibility, comparing samples of standard with similar materials stored similarly. The risk is parallel degradation. Another option, real-time stability testing, compares samples at normal storage temperatures with those at ultralow temperatures. This method reflects the actual degradation processes occurring at storage temperature. “For these two methods, changes are likely to be small, so their detection may require many assays,” Dr. Robinson explains.

“By their nature, proteins are more liable to decompose than small molecules,” says Stefan Bassarab, Ph.D., director of pharmaceutical development at Boehringer Ingelheim (www.boehringeringelheim.com), but newer formulations let you slow degradation.

New Formulations

“It is important to develop dosage forms that offer new ways to conserve the stability of proteins,” he adds. Emerging biodegradable polymers are one approach. Differential scanning calorimetry, to be discussed at the Prague meeting, can be used to examine the tendency of proteins to unfold at various temperatures and in different formulations to help researchers identify the most stable formulations.

“Biologics differ considerably in how they degrade,” emphasizes Johnson Varghese, Ph.D., director of process development at Amgen (www.amgen.com). “Therefore, it takes a variety of biochemical, biophysical, and biological methods that can be resource intensive and time consuming to identify the degradation pathways. In some cases, it takes a considerable amount of effort to identify critical degradants” such as immunogenic aggregates. Other challenges involve finding the right stress conditions for inducing meaningful degradations for a given protein.

Stressed stability testing is joining the collection of regular and accelerated stability testing for biologics. Because of the complexity of most biologics, Dr. Varghese says, “using stressed stability testing to evaluate the quality of the product can ensure that the product going to market is comparable to the clinical materials used to evaluate safety and efficacy.”

Stress Stability Testing

The test’s significant value during the development and commercialization of new or improved biologics comes in the form of information about the molecular structure, protease contamination, and primary degradation pathways and helps researchers “validate the stability-indicating nature of an assay, indicating analytical methods for future routine use,” Dr. Varghese notes. With statistical tools, it can also evaluate comparability of products manufactured at different scales or by slightly different processes.

“Stressed stability testing can afford insight within one or two weeks regarding the quality of the product,” Dr. Varghese says, “although identifying the exact structure and relative potencies of the break-down products may take much longer.” Regulatory agencies are requesting them as part of the comparability package, so they should be performed early and included in the chemistry, manufacturing, and control package.

At the Prague meeting, Timothy Schofield, senior director of nonclinical statistics, Merck Research Laboratories (www.merck.com), is discussing comparison of stability for biologics. Practical issues include the resources and time lines necessary to perform the studies, but quality is even more important when comparing new and old process materials and their effects on shelf life, he says.

Risks are mitigated by study design and analysis, he reports, with the goal of identifying changes in stability that may be linked to changes that may affect patients or shelf life.

Long-term, single-lot stability testing against specifications is standard, but “it doesn’t adequately address patient risk when a new material goes to market,” he says. “Such a study has little ability to detect a change, because there’s too little data to make a risk-based decision.

“We’re developing strategies, with an eye toward mitigating risk, to identify adaptive situations and define acceptable changes that are aligned with patient safety and efficacy,” Schofield says. Accelerated testing is one of those strategies. As an adjunct to long-term testing, accelerated testing gives an early read that is usually reliable, and it is actionable, giving an early indicator in a complex field in which many practices aren’t standard.

“The FDA,” he notes, “has said accelerated testing is reasonable.” Broaching the topic in Europe is one of the reasons Schofield is speaking at the Prague conference.

The challenges for mAbs are comparable to those of other biologics. “Even though they are bigger in molecular weight than blood-related biologics, recombinant hormones, growth factors, and other biologics, they are well-structured, more predictable, and relatively stable,” notes Hui Zhao, Ph.D., lab head of biotechnology development analytical R&D at Novartis Pharmaceuticals (www.novartis.com). For antibody drugs, it’s particularly important to understand their physical and chemical characteristics early in their development, she emphasizes.

Antibody Challenges

In her work to identify mAbs’ degradation pathways, Dr. Zhao says, “It is equally important to conduct degradation studies, identify susceptibility of the molecule to various degradation pathways, and select appropriate analytical methods to monitor changes that are likely to prohibit successful development.” She recommends substituting susceptible sequences before entering late-stage development.

“Degradation pathways vary significantly among various mAbs,” she says, “so changing a single amino acid or oxidation of some residues can dramatically alter the biological activities of molecules.

“For example, oxidation of methionine residues in the same variable region occurs only in some antibodies. Only specific tryptophan residues in particular sites of the antibodies are oxidized under the same photodegradation conditions. Therefore these changes are highly dependent on the structure of the molecules.”

The degradation pathways for formulation and nonformulation conditions are similar for mAbs. Forced degradation studies under nonformation conditions can identify degradation pathways and help select stability-indicating methods, she adds.

Efficacy of the molecule is affected by partial degradation, and the “degradation products from forced studies can be used for subsequent validation of stability-indicating methods and setting up specification of the product-related impurities. Degradation of some residues that affect potency including deamidation of asparagines or isomerization of an aspartic acid might occur only under formulated conditions. Such small changes could result in major changes in the overall structure of the antibodies. Therefore, it is necessary to perform degradation studies under both formulated and nonformulated conditions.”

Protein Aggregates

Detecting protein aggregates is particularly important because of the potential not only of harming patients when aggregates are undetected but also of contributing to the failure of many potentially good medications in clinical trials, says Tudor Arvinte, Ph.D., chairman and CEO of Therapeomic.

Formulations for peptides and proteins should be based on a detailed understanding of the physiochemical properties of the molecule “including structure, conformation, chemical degradation, and aggregation studies in aqueous solutions at different pH values, temperatures, buffers, and with different ingredients,” notes Dr. Arvinte, also a professor at the University of Geneva. The reason, he explains, is that differences all may affect aggregation, and “proteins behave differently when clumped together. Loose aggregates are more challenging to detect.

“For highly concentrated protein formulations, the detection of aggregates is challenging and the analysis is performed on dilutions,” Dr. Arvinte says. Traditional dilutions, after 100–200 times, may reveal mainly monomers “and you’ll think, wrongly, you’re safe,” he says.

Genentech’s (www.gene.com) Herceptin® may be a case in point. The label mandates dilution in a 0.9% sodium chloride solution but never in a 5% dextrose solution. “Although the EMEA summary of product characteristics states that dilution of Herceptin in dextrose results in protein aggregation, no standard aggregation methods can detect easily these aggregates.” To do so, new analytical methods were used.

Dr. Arvinte recommends methods that do not require dilution such as free-flow fractionation and fluorescence using the hydrophobic Nile red stain. The latter reveals aggregates that may not be detected by other methods and in solutions that appear clear to the naked eye. Light-scattering methods and fluorescence are particularly effective at analyzing protein aggregation or binding to surfaces for formulations with low protein concentrations.

The goal, he says, “is to be able to measure high and low concentrations of aggregates in formulations before they are given to patients.”