The 1999 World Congress of Pharmacy and Pharmaceutical Sciences - the 59th international congress of the International Pharmaceutical Federation (FIP) - took place in Barcelona, Spain, from September 5 to 10. Our coverage continues with reports of two symposia on international aspects of regulatory affairs (Harmonising specifications for chemical and biological quality characteristics and Harmonising treatment of marketing authorisation variations), both held on September 7
Opening the first symposium, on the harmonisation of specifications for chemical and biological quality characteristics, the chairman, Professor JOSEP TORRANT (Spain), reminded the audience that the series of International Conferences on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Humans (ICH) was nearing a global agreement (the so-called "Step 4") for setting drug product specifications. These were critical quality standards that the manufacturer justified to regulatory authorities as conditions of approval. However, in practice, the nature or origin of the product strongly influenced the approach to setting specifications.
Professor HELGA MÜLLER (Central Pharmaceutical Laboratory, Germany) examined the differences underlying the development of guidelines in ICH group Q6A for world-wide specifications for chemical substances. The general concept was to establish an agreed minimum of data acceptable to all national agencies. For example, Europe placed a traditional emphasis on tight product-release specifications, while the United States and Japan relied on limits derived from shelf-life studies. ICH had compromised by requiring the latter, while leaving the former for in-house testing.
ICH group Q6A recognised in-process controls as useful for checking key parameters critical to a manufacturing process, but these could be omitted if release tests were satisfied. Europeans were ready to use "skip" tests in place of full periodic testing if this could be justified by the experience of testing during product development (eg, if levels of residual solvents had been shown to be consistently low). Similarly, "parametric release" could be justified to the appropriate agency as an acceptable alternative to routine testing to the limits of a release specification.
Professor Müller considered that a well described and controlled drug substance (particularly with rigorous testing of impurities arising in synthesis) reduced the requirements expected in a drug product specification while retaining the need to control prospective degradation artifacts.
On the prospects for harmonising compendial specifications, Professor Müller said that ICH had considered "general topics" such as uniformity of content, disintegration and dissolution testing, and had "essentially harmonised" requirements for sterility, mineral residue, bacterial endotoxins, preservative efficacy, particulate matter and clarity of solution. She believed that the US Pharmacopeia (USP) and the European Pharmacopoeia (PhEur) had reached a common position on interchangeability and nominally equivalent requirements in their respective pharmacopoeias, whereas Japan cross-referred to Europe or US. All still needed to agree on choice of media and acceptance criteria for dissolution studies and uniformity of content.
Dr BRENDAN HUGHES (Glaxo-Wellcome, UK) reviewed the achievement of ICH group Q6B in relation to common specifications for biological substances and biotechnology products. Given the breadth of biotechnology, the guidelines demonstrated general principles with a gradual harmonisation that should reduce the load of testing in developing products that could be satisfactorily characterised physico-chemically.
The European Medicines Evaluation Agency recognised the guidelines as from September, 1999. For biotechnology products, the scope went beyond the Q6A's chemical substance guidelines in requiring conformance at "other stages of manufacture", with specification limits for critical parameters rather than (the more complex) full characterisation of every batch. There were expectations at those stages that were accessible to high purification and characterisation. The guideline for the marketed product (rather than clinical trial material) still relied heavily on tests during development, especially immunological tests. The products for Q6B tests were proteins, polypeptides and cell cultures but not antibiotics and vitamins (for which Q6A applied). Vaccines and DNA products were outside the scope of both.
Characterisation of the biotechnology product was a major part of the guidance. Agencies recognised that there would be heterogeneity with a series of related products.
Assessment of biological activity was another essential element. Potency had to be defined, preferably related to a national reference substance, but for a novel substance a fully characterised in-house standard might be acceptable. An agency would accept a wide range in the potency if sufficient physicochemical data could be provided, ideally supported by separate bioassay and quantitative assessment of protein.
Dr Hughes did not expect ICH to recommend specific limits for impurities. The agency should judge on the merits of the product, recognising the molecular complexity. Results would be method dependent and it was recommended to combine several methods. Variation could also be expected in the product of synthesis - eg, variable degrees of glycosylation of peptides.
Dr Hughes distinguished between product-related impurities, which could affect bioactivity, and process-related, which might affect final structure. Both collective and individual limits would be set with some allowance for contamination - but separate guidelines on viral contaminants should be followed. As with small molecule specifications, standards needed to be calibrated and analytical methods validated but one major difference with biologicals was the confidence to be drawn from in-process controls, with limits assigned from development experience. There would also be special biological requirements for starting materials. He concluded that a developer should justify each stage within an overall control strategy.
Dr CHRIS POTTER (Astra-Zeneca, UK) considered how Q6A guidance for solid oral drug specifications could be extended to setting specifications for topical semi-solid products and transdermal delivery systems (TDDS) for treatment of unbroken skin and special dressings for open wounds. He contrasted PhEur definitions of single-phase "ointments", multi-phase "creams", liquid "gels" and semi-solid "pastes", with the USP's definitions for creams (narrower) and ointments (wider). However, there was common ground between PhEur and USP on TDDS with diffusion-controlled delivery of an active drug through a skin barrier into systemic circulation.
The guidance of Q6A on universal test criteria - which it was hoped would be adopted at Step 4 this autumn - could mostly be applied to single- and multi-dose topical products and to TDDS. Tests that were not relevant were inorganic impurities, extractables, redispersibility and reconstitution, and specific gravity. Additional tests, not harmonised by Q6A, included minimum fill, homogeneity for multi-dose topicals, and dissolution studies.
The PhEur and USP had different apparatus and conditions for dissolution studies, and also different dose limit definitions. For nicotine patches, the USP offered four choices of apparatus and rotation speed. Like Professor Müller, Dr Potter picked out from Q6A the opportunities for periodic ("skip") testing and, ultimately, parametric release, and he noted potential confusion in the concept of in-process tests. He hoped for harmonisation, or mutual recognition, in dose uniformity, preservative efficacy and dissolution tests. However, major barriers to world-wide specifications remained in the lack of pharmacopoeial harmonisation, the continuing cross-Atlantic debate on release limits v shelf-life, the proliferation of dissolution procedures, and the different expression of limits.
Addressing specifications for metered-dose inhalers (MDIs), Dr BRIAN ROGERS (Food and Drug Administration, US) said that MDIs were an increasingly popular form of medication. Their unique features included their special formulations and dispenser engineering, which were critical to the inhalation route, and the importance of maintaining a uniform dosage. They had special chemicophysical properties and needed relevant control parameters. US conventional monograph specifications for excipients might have to be supplemented to ensure lot-to-lot reproducibility.
Prior to critical clinical trial studies, comprehensive ex vivo performance should be established and reconciled with appropriate validated tests and corresponding acceptance criteria. Of the extensive lists of recommended test parameters, four were "critical": moisture content, content uniformity throughout product life, amount and toxicity of leachables, and precision "valve geometry". The regulatory perspective began in early development and clinical trial stage, and the ultimate application should reflect all the wealth of experience acquired during product development.
Finally, Dr JOSÉ FÁBREGAS (Barcelona, Spain) noted that there was less regulation for starting materials that were stable if isolated and analytically well defined. Guidelines for applications to the Committee on Proprietary Medicinal Products in Europe differed from those of the FDA, as now recommended by ICH. There were differences of terminology (eg, "intermediate", "key", "pivotal" and "final" preparative stages) and also differences of emphasis (eg, ICH guidelines of July, 1999, concentrated on structure and chemical properties).
Using the synthesis of zomepirac as an example, Dr Fábregas contrasted the appropriate specifications for identity, assay and impurities of main starting materials, with limits for unreacted solvents and isolated intermediates. He had two major concerns with the quality of starting materials: a possible secondary synthesis might lead to a by-product relatively similar to the desired active ingredient; and other interference with good manufacturing practice. Another issue was quality guarantees from suppliers of starting materials and the traceability of their quality records.