A full-day symposium on the analysis of biopharmaceuticals took place on September 16. Organised in association with the Joint Pharmaceutical Analysis Group, it was chaired by Dr H. M. Hill
The symposium started with an address on the harmonisation that is needed, and that should be achieved in the near future, all being well, between the various pharmacopoeias with regard to the analysis of biopharmaceuticals.
Dr ADRIAN BRISTOW (National Institute for Biological Standards and Control, Potters Bar) explained that biological medicines were usually defined to be those that could not be adequately characterised by physicochemical methods alone, but required the use of some form of bioassay. Over the past 15 years or so, much progress had been made — for example, by recombinant DNA technology — in synthesising the active principles of what had previously been somewhat incompletely purified blood or tissue extracts, so that they attained the status of being a single chemical entity. In such cases, bioassays, which were expensive and laborious, could conceivably be dropped, to be replaced by simpler chemical assay methods. The speaker identified a number of divergences of treatment in the pharmacopoeias, and described a document currently being drafted by the European Pharmacopoeia that was a guide to the elaboration of monographs for biologicals. Completion and implementation of this would go a long way towards producing uniformity.
An extension of this theme was given by Dr L. TSANG (Medicines Control Agency), who reviewed the various European Community directives that had been issued over the past couple of decades, and their effects on manufacturing practice and analytical control.
The testing of blood for viral contaminants was essential to safety, said Dr JOHN SALDANHA (National Institute for Biological Standards and Control). The major contaminating organisms were those of hepatitis A, B and C, human parvovirus, and HIV.
Precautions against viral transmission included selection of donors, screening of donated blood and viral inactivation/removal during manufacture of blood products. But, despite all this, there could still be virus transmission, not least because of the “window period” during which blood from an infected donor might be antibody negative but still viraemic.
The latest technique for the detection of viral contamination was nucleic acid amplification, which was sensitive enough to reduce the window period by about 50 per cent. In the worst case, hepatitis C, the reduction was from 82 days to 59, which was still a long enough time for a virus to slip through the net. In Germany, there had been a requirement to test erythrocyte concentrate by this method since April, and the European Committee for Proprietary Medicinal Products had required it since July for the testing of plasma pools.
This type of assay, of which the polymerase chain reaction was an example, was extremely sensitive: indeed, a single molecule might be replicated a million times, so that cross-contamination and false positives became a problem, as also did false negatives that might occur as a result of small variations that led to suboptimal amplification.
As with any test intended for pharmacopoeial use, standardisation was essential. In 1995, the same sample had been sent to 15 laboratories with wildly divergent results. Matters had improved greatly since then. Working standard reagents had been established by the NIBSC for the hepatitis A, B and C, human parvovirus, and HIV organisms, and there was now a World Health Organisation international standard for hepatitis C, against which several working reagents had been calibrated. There was also an international unit, which should supersede the variety of units that had been current up to now, and should help in the standardisation process.
Discussing the use of mass spectrometry in elucidating the structure of proteins and glycoproteins, Dr FIONA M. GREER (M-Scan Ltd, Ascot) said that the two conventional methods of determining the primary structure of a protein were Edman degradation with automated gas phase sequencing and deduction from the DNA base sequence. Both methods had their disadvantages, in particular their inability to deal with post-translational modifications.
Over the past 10 years, advances in mass-spectrometric technology had made mass spectrometry an indispensable tool, since it could determine the identity of blocked N-terminal residues and the presence or absence of C-terminal residues, and it could handle glycosylation and the analysis of disulphide bridges.
Great strides forward had been made in the generation of ions, which was the start of the mass-spectrometric process. New techniques that had contributed to the production of larger ions than had previously been possible included fast atom bombardment (FAB), in which a beam of atoms was directed at the specimen, electrospray (ES), in which a solution was sprayed through an orifice into an intense electric field gradient, and matrix-assisted laser desorption/ionisation (MALDI), in which the specimen, on a metal plate, was heated by a laser beam, possibly with the addition of a chromophore to enhance the effect of the laser.
With the introduction of time-of-flight (TOF) measurement, six- or seven-figure precision was now attainable on ion masses of 20 kDa or more. Dr Greer gave a number of numerical examples, in which structures had been determined by finding the arithmetical differences between the masses of the different ions appearing in the spectrum.
Dr ASHVIN PATEL (Glaxo Wellcome) gave an account of the development and optimisation of a biosensor method for determining the affinity binding of a monoclonal antibody. The equipment was a Biacore biosensor, which operated on the principle of attaching an antigen to a dextran substrate to immobilise it. The dextran surface was then viewed by polarised light through a prism, which gave shadow bands in the reflection. Antibody was then added to form a complex with the antigen. This changed the surface film and caused a shift in the position of the shadow bands. The injection of a suitable buffer then removed the antibody and regenerated the biosensor surface.
The United States Food and Drug Administration recommended that, during the development and validation of an assay method, all the variables likely to affect the test were to be identified, measured, understood and controlled. This had been achieved by factorial experimental planning, enabling a response surface to be plotted that gave the flow rate and buffer conditions that were optimal in terms of the accuracy of the assay.
The final paper of the symposium was presented by Dr LANCE SMALLSHAW (Eli Lilly, Speke) who, over a 17-year period in the analytical section of the company, had followed the difficulties attending the analytical control of recombinant human growth hormone, Humatrope. A change in the EP monograph had sparked off the company’s “analytical harmonisation project”, which involved the setting up of a team of 18 people to examine procedures at Lilly’s two manufacturing locations, Speke and Indianapolis. This had been an expensive exercise, but it had to be viewed in the context that somatropin (of rDNA origin) for injection was valued at £20,000 a gram.
A comparability document had been drawn up for the UK and US procedures, plus an expert report, and all the analytical methods had been rigorously re-examined. Where divergences had been found, they had successfully been eliminated.