State-of-the-Art vs. Tried and Trusted

Development of biopharm products is driven by quality, time to market, and cost, and it is these factors that have forced an evolution to state-of-the-art analytical test methods that have gradually replaced old tried-and-tested approaches. There has always been a healthy conflict between the need to progress and improve methods, balanced against the greater certainty of traditional methods. So as time goes by and confidence is gained, the state of the art becomes the next generation of tried and tested. But the progress is not always either continuous or straightforward.

This article illustrates some factors and pitfalls that influence progress from tried-and-tested methods to state-of-the-art ones. After the principle factors of quality assessment are summarized, examples are provided from the author's experiences during the past decade where different driving factors have forced replacement of old methods with new. And as a demonstration that new is not necessarily better, in one example the problems encountered with the new method forced a return to the old.

Biomanufacturing Revolves Around Quality and Testing

Reliable quality data is the focus of all regulatory investigation and licensing. Analytical testing is necessary during biomanufacturing to demonstrate high levels of control and quality, thus providing confidence that the final active product is consistently pure, efficacious, and safe. The scope of analytical testing encompasses every stage of manufacture, from master cell bank to final batch release testing and stability studies. This testing is not limited to actual product testing, as it is also essential to monitor the performance and quality of the manufacturing facility itself.

A potentially huge amount of analytical data is produced during routine biomanufacture, typically generated by numerous orthogonal technologies. Analytical development is, therefore, essential to ensure that the information provided is usable and focused. Fundamentally this involves the application of the most suitable analytical techniques, in terms of sensitivity, specificity, and regulatory acceptability. Many of these techniques are now well-established and have been optimized for their application; however, their development for new applications is continuing.

Figure 1a. Structural Features of Example Biopharmaceuticals.

So What Defines Product Quality

Biological medicinal products form a broad range of product types including nucleic acid, proteins (Figure 1), viruses, and cell therapies. The development of biological materials as therapeutic agents requires the comprehensive assessment of quality with regard to regulatory guidelines such as those of the International Conference on Harmonisation (ICH) (Figure 2). The quality of a biotech product is defined by both the product itself, which should be both characterized and consistent, and the manufacturing process, which must be both validated and reproducible.

Figure 2. ICH Guidelines Applicable to Manufacture of Biotech Products16

The assessment of product quality is determined using several criteria including:

• Physico-chemical properties

• Biological activity

• Immunochemical properties

• Purity, impurities, and contaminants

• Quantity

• Identity

• Safety

A well-characterized product¹ is critical during development, both to establish batch-to-batch consistency, and to establish analytical tests and specifications for process control and product release (Figure 3). In addition, the assessment of product stability relies on the adequate product characterization prior to and during a stability testing program.

Figure 3. Characterization Testing as the Basis for Choosing Suitable Quality Control Tests

Current Trends in Manufacturing Philosophy

Test methodologies have certainly been influenced by the latest manufacturing philosophy. The US Food and Drug Administration (FDA) has stated, "Quality cannot be tested into products; it should be built-in or should be by design."² This concept of building quality into a medicinal product is reliant upon having a thorough understanding of the manufacturing process, as from this, greater control may be achieved. Process analytical technology (PAT) is a system that enables greater understanding of the manufacturing process by providing a mechanism to design, analyze, and control manufacturing through taking measurements of critical quality parameters during processing. This in turn provides a procedure to ensure acceptable end-product quality at processing completion.

The PAT system offers substantial benefits, yet several challenges and threats must be overcome if it is to be successful in biomanufacturing. These include:

• Innovation to develop analytical tools that assess critical attributes presently unmonitored

• Investment required to implement novel technologies

• Increased complexity of required data-retention for huge data volumes generated by continuous monitoring

• Likelihood of stringent acceptance specifications due to increased data

• Regulatory uncertainty and need for more guidance

The Biogeneric Debate

Regulatory authorities are debating the extent to which product similarity can be trusted for biogeneric products, and they have indicated that new testing regimes will eventually be required on a case-by-case basis. This will undoubtedly influence the development of new and improved analytical methods.

In the key US market, all drugs are governed by the Federal Food, Drug, and Cosmetic Act (FDCA). To gain FDCA approval for generic pharmaceuticals, an applicant must file an Abbreviated New Drug Application (ANDA) scientifically demonstrating that the product is bioequivalent and that it is comparable to an innovator drug in dosage form, strength, route of administration, quality, performance, and intended use. The approval process for generics in Europe is similar to that in the US. There, the procedure is overseen by the European Agency for the Evaluation of Medicinal Products (EMEA), is rapid, and does not require clinical trials if bioequivalence is demonstrated.

By comparison, there is presently no similar pathway for biogenerics either in the US or Europe. In the US, most protein therapeutics are classified as biologics and are governed by the Public Health Service Act (PHSA). The PHSA provides a Biologics License Application (BLA) for new biologic marketing approval, but it lacks an abbreviated pathway for generic biologics. In Europe the process for biologics has not been finalized, but the EMEA recognizes the complexity of such molecules; therefore, under proposed changes, biosimilar products may, or may not, be molecular copies of the original product. The products must, however, depend on the same mechanism of action and be intended for use in the same therapeutic indication. It is still unclear if in Europe a biogeneric should be registered through the abbreviated procedure applicable for well-defined traditional generics or undergo a full registration process.³

The EMEA has recently issued guidelines (currently under consultation)^{4,5, 6,7,8,9} for producing biogeneric medicines (referred to as "biosimilar"). FDA, meanwhile, is continuing to discuss its position on biogenerics, and draft guidelines are expected later this year. The debate on biosimilar products is still continuing; however, it is clear that the analytical tools need to be available to prove that a generic product is essentially similar to the reference product. Discussions with regulatory agencies can lead to new state-of-the-art methods being adopted, or conversely, to a retrenchment to more traditional methods, as illustrated in the following examples.

An Example in Which State of the Art is Clearly Better

State-of-the-art polymerase chain reaction (PCR) has gradually replaced Southern blotting because Southern blotting is too slow and cumbersome. During the late development phase of a complex recombinant protein vaccine in the late 1990s, it was noted that the stability of one of the genetic elements of the cell line was questionable during production. Rather than redefine the production method it was agreed that the company would provide a routine analysis for gene copy number. The classical method for detection of the genes was Southern blotting.¹⁰ However, it became apparent that this method had serious problems as a quality control test: it was very time-consuming, variable, and cumbersome.

The most promising technique found to substitute for Southern blotting was PCR, the basic method invented in 1983 by Kary Mullis of the Cetus Corporation. PCR is a process where DNA can be artificially multiplied through repeated cycles of duplication driven by an enzyme called DNA polymerase.

The classical PCR test has advantages of specificity and time over Southern blotting; however, it has limitations. The major drawback is the lack of accurate quantitative information due to amplification efficiency. In the later cycles of the PCR, the amplification products are formed in a nonexponential fashion at an unknown reaction rate, and so the link to the initial quantity in the sample is lost.¹¹

DNA quantification based on conventional PCR relies on endpoint measurements to achieve the maximum sensitivity. At this stage the reaction has gone beyond the exponential phase, and the resulting correlation between the final product concentration and the number of initial target molecules is therefore limited.

In 1993 Higuchi et al published an analysis of PCR kinetics; researchers had set up a system to detect PCR products as they accumulated "real time."¹² By 1997 the first high-throughput real-time thermal cycler was on the market; it has revolutionized the detection and quantification of nucleic acids in many areas of product development.

Table 1. Comparison of Main Validation Criteria Between Southern Hybridization and QPCR

In an evaluation of Southern hybridization and quantitative PCR (QPCR), a comparison of the validation data for each assay indicated that QPCR is a more robust and accurate method, provided that all controls are in place (Table 1). The following main advantages of quantitative PCR over Southern blotting outweighed the disadvantages of capital and running costs:

• Assays highly specific for each construct

• Simple 96-well plate format provides high sample throughput

• Rapid assay compared to Southern blotting (one day opposed to one to two weeks)

• Small amounts of DNA are required (<100 ng per assay opposed to 10 _g for the hybridization assay)

• Assay sensitivity is < one copy per cell

• High precision and accuracy

• Safety: no radio-labeling of probes

• No cumbersome manipulation of agarose gels and blots

Assay development time is an important factor in an overall development program. The PCR assay took approximately nine months and two members of staff to specify and purchase the equipment, and then to develop and validate the test to the ultimate satisfaction of the authorities. The regulatory bodies then required a large volume of data to compare the two techniques before they were satisfied that the PCR method could replace the routine Southern blotting method. They also required the use of both techniques during product release of a number of manufactured batches. As more and more applications now contain PCR-based measurements, the regulatory authorities are accepting PCR as the de facto standard.

An expample in Which State of The Art Created Confusion

Sometimes the complexity of state-of-the-art methods and their accompanying data lead to inconsistencies rather than increased accuracy. Unless there are clear benefits from the new tests, regulatory authorities can force a retrenchment to the old tried and trusted ways. This was the case with a carbohydrate analysis of the same recombinant vaccine development program conducted in the late 1990s.

The oligosaccharide mapping (N-linked) was performed by fast-atom-bombardment (FAB) mass spectroscopy (MS) and matrix-assisted, laser-desorption/ionization – time of flight (MALDI-TOF) MS as a characterization test on the active substance. After trypsinization of the sample, the N-glycans were released by N-deglycosylation with subsequent reverse-phase high-performance liquid chromatography (rpHPLC) purification. The purified oligosaccharides were permethylated and analyzed by FAB-MS and MALDI-TOF MS, respectively.

Additionally, carbohydrate analysis and total carbohydrate analysis were performed by gas chromatography (GC) MS as a release test on the drug substance. The glycosidic bonds were cleaved by methanolysis and the released monosaccharides modified to give volatile trimethylsilyl derivatives, which were subsequently separated by gas chromatography and identified by mass spectrometric analysis. Identification of individual methylglycosides was based on retention time and mass spectrum by comparison with the standard mixture. Total carbohydrate was derived by summing the individual monomers.

The proposed mass spectrometry method to monitor glycosylation was considered to be cumbersome and unsuitable, as it lacked sensitivity and accuracy to monitor batch-to-batch consistency of glycosylation. The company was asked to provide convincing ([semi]-quantitative) data with regard to the batch-to-batch consistency of the glycosylation pattern (e.g., by fluorophore-assisted carbohydrate electrophoresis [FACE]).

FACE, using the enzymic release of oligosaccharides followed by electrophoresis, was investigated. An early semi-quantitative version of this showed good correlation between the relative quantities of the oligosaccharide species released from the samples and so provided evidence for a consistent glycosylation pattern between batches, both in terms of relative amounts and species present. Although this gel-based method could not differentiate between oligosaccharide species with the accuracy of MS, it was considered by the regulators to be less of a risk than adopting the then-experimental MS technique.

Eventually, however, state of the art succeeds. Even if a state-of-the-art technique such as MS is rejected in favor of an old tried-and-trusted method in some situations, it can still prevail when a change is forced by driving factors, such as the power of MS to separate similar protein isoforms, which cannot easily be done with accuracy by any other method. ¹³

The Evolution of Protein-Characterization Techniques

The ability to separate molecules based on size and charge was first described in 1912 by JJ Thompson. Despite years of intense development of MS methodology and equipment, the goal of analyzing large macromolecules remained elusive for over 70 years.¹⁴

Before 1982, several tools were available to characterize natural proteins physico-chemically. These were:

• Edman degradation

• UV and visible spectroscopic methods

• Protein content tools (such as the Bradford method)

• Constituent chemical content (e.g., nitrogen)

• Electrophoretic analysis (e.g., sodium dodecyl sulfate polyacrylamide gel electrophoresis [SDS-PAGE])

The emphasis on the "process defines product" attitude during the early development of biological products was due to the relative inability to fully characterize protein products to ensure safety and efficacy. The big breakthrough in protein characterization of the 1980s was the advent of high-performance liquid chromatography (HPLC). HPLC provides high peak resolution and can be demonstrated by the separation of species-specific insulins, many of which differ by a single amino acid.¹⁵ HPLC was employed for the characterization of insulin as the first recombinant protein therapeutic to be approved. HPLC is now a mainstay of the protein characterization toolbox.

Similarly with the advent of better detector technology, MS was able to resolve higher molecular weights. This allowed increased assurance of product quality and a better understanding of the relationship between structure and function. For instance, the total molecular weight of proteins up to approximately 500 kDa can be determined by MS with a far greater mass accuracy than by using other methods such as SDS-PAGE and size-exclusion chromatography. Accuracy and resolution are further increased as the mass of the analyte decreases, such as when proteins are fragmented to peptides during peptide mass fingerprint. The exceptional mass accuracy of MS permits the characterization of major post-translational modifications, for example disulphide bonding, glycosylation, and lipidation.

Until recently MS was rarely used as an in-process or release test, probably because MS has been considered a highly specialized procedure that is assumed to be unsuitable for routine manufacturing. However, the rise of proteomics has led to mass spectrometers that are high-throughput and robust, and therefore, ultimately more cost-effective and easier to use. So it may yet transpire that validated MS will become more commonly used as an in-process test. This may actually be a necessity in certain cases where purification involves the separation of isoforms that may only be differentiated by a technique with high mass accuracy.

Conclusion

Factors such as quality, time to market, and cost have forced an evolution of state-of-the-art analytical test methods. In addition, changes in the regulatory framework, such as PAT and biogenerics, can also drive innovation in analytical development. However, putting these techniques into routine use can be challenging. Some powerful state-of-the-art techniques are PCR and MS. (The uses for these techniques and others in protein characterization can be found in Table 2.) New techniques require evaluation and development to ensure that the data they provide is equivalent to that of more accepted and traditional methods, but nothing can replace good scientific rationale and careful experimentation and testing.

Table 2. Application of Assay Techniques

Anita Bate, Ph.D., is a team member at Eden Biopharm Ltd, D5, Stanlaw Abbey Business Centre, Dover Drive, Ellesmere Port, Cheshire, CH65 9BF, UK, 44 (0) 151.356.5632, Fax: 44 (0) 151.356.5633, anita.bate@edenbiopharm.co.uk

References

1. Zoon K, Gornik R. Definition of a well characterized biotechnology product. Dev Biol Stand. 1998;96:191-197.

2. US Food and Drug Administration. Guidance for Industry, PAT — A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance. Rockville, MD: Office of Training and Communication, Division of Drug Information, HFD-240, Center for Drug Evaluation and Research, Food and Drug Administration; September 2004. Available at: http://www.fda.gov/cder/guidance/6419fnl.htm.

3. EMEA Human Medicines Evaluation Unit, Committee for Proprietary Medicinal Products. Note for Guidance on Development Pharmaceutics. London: EMEA; 1998. (EU Directive 65/65/EEC Article 4.8, medicinal product marketing license directives.) Available at: http://www.emea.eu.int/pdfs/human/qwp/015596en.pdf.

4. EMEA Human Medicines Evaluation Unit, Committee for Proprietary Medicinal Products. Draft Guideline on Similar Biological Medicinal Products. London: EMEA; 2004. Available at: http://www.emea.eu.int/pdfs/human/biosimilar/043704en.pdf.

5. EMEA Human Medicines Evaluation Unit, Committee for Proprietary Medicinal Products. Draft Guideline on Similar Biological Medicinal Products Containing Biotechnology-Derived Proteins as Active Substance: Quality Issues. London: EMEA; 2005.

6. EMEA Human Medicines Evaluation Unit, Committee for Proprietary Medicinal Products. Similar Biological Medicinal Products Containing Recombinant Human Growth Hormone (Annex to Guideline). London: EMEA; 2004.

7. EMEA Human Medicines Evaluation Unit, Committee for Proprietary Medicinal Products. Similar Biological Medicinal Products Containing Recombinant Human Erythropoietin (Annex to Guideline). London: EMEA; 2004.

8. EMEA Human Medicines Evaluation Unit, Committee for Proprietary Medicinal Products. Similar Biological Medicinal Products Containing Recombinant Granulocyte-Colony Stimulation (Annex to Guideline). London: EMEA; 2004.

9. EMEA Human Medicines Evaluation Unit, Committee for Proprietary Medicinal Products. Similar Biological Medicinal Products Containing Recombinant Human Insulin (Annex to Guideline). London: EMEA; 2004.

10. Southern EM. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol. 1975;98:503.

11. Weighart F. Quantitative PCR for the Detection of GMOs. European Commission Joint Research Centre Workshop: The Analysis of Food Samples for the Presence of Genetically Modified Organisms, Session 10. World Health Organization: 2004. Available at: http://gmotraining.jrc.it/docs/Session10.pdf.

12. Higuchi R, Fockler C, Dollinger G, Watson R. Kinetic PCR: Real-time monitoring of DNA amplification reactions. Biotechnology. 1993;11:1026-1030.

13. Gevaert K, Vandekerckhove J. Protein identification methods in proteomics. Electrophoresis. 2000; 21(6):1145-1154.

14. Markides K, Gräslund A. (2002) Advanced information on the Nobel Prize in Chemistry 2002. Stockholm; The Royal Swedish Academy of Sciences: 2002. Available at: http://nobelprize.org/chemistry/laureates/2002/chemadv02.pdf.

15. Rivier J, McClintock R. Reverse phase high-performance liquid chromatography of insulins from different species. J Chrom. 1983;268:112-119.

16. Seamon KB. Specifications for biotechnology derived protein drugs. Curr Opin Biotech. 1998;9:319-325.