Return to PJ Online Home Page
The Pharmaceutical Journal Vol 264 No 7082 p226-227
February 5, 2000 Forum

JPAG/PSG/IPG

Control of fermentation processes

The control of fermentation processes was the topic of a recent symposium organised by the Joint Pharmaceutical Analysis Group in association with the Royal Pharmaceutical Society's Pharmaceutical Sciences Group and Industrial Pharmacists Groups

Even the smallest change to a pharmaceutical fermentation process needs thorough investigation, speakers warned the symposium's participants. A new strain of a micro-organism or a minor process change could lead to vastly increased concentrations of an impurity or the introduction of an impurity that had not previously been experienced. In one well-known incident, a number of deaths had been caused by potent new contaminants in a product that continued to meet its original purity specification.

Industrial overview

Giving an industrial overview of fermentation controls, Dr MATTHEW DUCHARS (Zeneca Life Science Molecules) said that they were needed both for optimisation - to understand critical control parameters (CCPs) and to develop best product formation conditions - and for reproducibility - to maintain tight limits on CCPs for robustness, to achieve consistent product quality and to stay within regulatory boundaries.
Dr Duchars demonstrated a cycle of feedback such that "input controlled output" and where results from analysis of output parameters, quality, quantity and productivity were fed back to influence what should go into the process in terms of, for example, nutritional and physical requirements. He also stressed the importance of established genetic control prior to fermentation.
Giving examples of the variety of testing and controls for different types of products and the response times required, Dr Duchars used three case studies. The first was continuous culture at 140,000 litres of Fusarium graminearum (Quorn). For a biomass product such as this, growth rate was linked to its branching morphology and the best control was found to be indirectly linked to carbon dioxide production, which could be monitored automatically on-line.
The second case study was a biotransformation of a chemical drug intermediate. The biotransformation used Neurospora crassa and was shown to be the only economic route to achieve the high stereoselectivity required to reduce a ketosulphone feedstock to the correct hydroxysulphone derivative. However, the keto starting material had been found to be toxic to the reaction and, while both materials could be monitored by high performance liquid chromatography off-line, on-line near-infrared spectroscopy was found to be more applicable to monitor the formation of the hydroxysulphone and hence control the addition of the more toxic ketosulphone.
The third case study was the production of a recombinant therapeutic protein in yeast. Because the protein required extensive purification downstream, it was not possible to test for product on-line or at-line. However a multifunctional experiment revealed that a pH suboptimal for growth allowed optimal product purification, and hence pH became the key control.
Dr Duchars finished by saying that while on-line controls were preferred, more complex products required individual approaches with the emphasis reverting to pre-process development to ensure product quality.

Regulatory overview

Providing a regulatory overview, Dr ADRIAN THOMAS (Medicines Control Agency) began by looking at the types of substances produced by fermentation. He categorised these as indirect gene products (such as antibiotics, vitamins and toxins) or direct gene products (such as hormones, monoclonal antibodies, proteins and antigens). Such products required approval under the mandatory centralised procedure for medicinal products or excipients developed by recombinant DNA technology (European Community regulation EEC 2309/93).
Because some more established processes were being updated post-approval by introduction of a biotechnology manufacturing step, Commission Regulation (EC) No 542/95, as amended by EC No 1069/98, came into force. This excluded proteins, food or food ingredients, but was aimed at capturing any changes that could result in a new impurity which would then need to be described, with a test procedure, in a European Pharmacopoeia certificate of suitability, which a new version of amended resolution AP-CSP(98)2 had come into force in May, 1998.
Dr Thomas then presented a series of issues for which there were significant safety concerns and which manufacturers needed to consider when applying for regulatory approval. The main concerns centred around ensuring freedom from adventitious viruses and transmissible animal spongiform encephalopathy agents, especially in sourcing raw materials and establishing cell banks. Having a validated cell bank that was shown to be stable and free from pathogenic agents was seen to be critical for any fermentation process.
In-process controls were also highlighted as a means of ensuring process reproducibility but the effects of changes on the process had to be revalidated. To emphasise the importance of such revalidation, Dr Thomas described an incident in 1989 when a new strain and minor process change in the production of l-tryptophan had resulted in over 30 deaths and hundreds of cases of eosinophilia-myalgia syndrome (EMS). The purity of the product had still met its original specification but some very potent new trace contaminants had not been removed. Subsequent work had identified these, and there was now a monograph in the European Pharmacopoeia with limits of 10ppm maximum for 1,1-ethylidene bis-tryptophan.

European perspective

Professor DEREK CALAM (president, European Pharmacopoeia Commission) also referred to the l-tryptophan/EMS incident, citing it as the issue that had exposed a need for a monograph, since even with a purity of 99.6 per cent, deaths had still occurred.
Professor Calam reviewed the regulatory processes from 1987 when a directive to pool expertise on recombinant technology had led in 1993 to the introduction of the centralised licensing procedure. In 1996 a special group of experts had been constituted to produce a monograph for products of fermentation. This had come into force in January, 1999 - one year ahead of schedule.
The monograph covered active or inactive pharmaceutical substances produced by controlled fermentation as indirect gene products and primary or secondary metabolites of micro-organisms whether or not modified by traditional technology or recombinant technology (eg, vitamins, amino acids, antibiotics, alkaloids or polysaccharides). Related to such products the key issues were manipulation of the organism, process control, change control and validation. Of particular importance were the history and characterisation of the micro-organism, controls over cell banking, raw materials, and in-process controls.
Change control received special attention because of the potential impact of change on the impurity profile - either the introduction of new impurities or increases in the levels of existing impurities. Replacement of a master cell bank required revalidation of critical steps and demonstration that no adverse changes to quality and safety of the product had occurred. If the micro-organism itself was modified or replaced, then close attention to the impurity profile was again required, with careful assessment of any changes in quantity or identity. Significant revalidation was essential after all such changes.
As a consequence of this monograph, the certificate of suitability was to cover only indirect gene products and the 1993 resolution for the centralised procedure had been clarified accordingly. Thus the scope of the certification procedure was such that it might apply to substances produced by fermentation as indirect gene products, which were metabolites of micro-organisms, irrespective of whether or not the micro-organisms had been modified by traditional procedures or recombinant DNA technology. The procedure would not be applicable for direct gene products (proteins) or substances obtained from human or animal tissues [Resolution AP-CSP (96)5 of the Council of Europe as modified in 1998]. Professor Calam was keen to emphasise that this monograph was a result of the Pharmacopoeia working rapidly and interactively to produce a solution acceptable to all.

Implications for animal medicines

Reviewing the implications for veterinary medicines, Mrs JANICE KIRBY-SMITH (Veterinary Medicines Directorate) said that the regulatory issues were very similar to those for human medicines. Veterinary products could be divided into two clear classes. Pharmaceuticals, which included medicinal feed additives, were controlled under Directive 81/851/EEC. But other feed additives were controlled under a separate Directive, 70/524/EEC.
The active substance might be isolated or not isolated (feed grade). For fermentation control the feed grade substances presented some additional issues to those for isolated active substances. Assay and tests for related substances on the active component were required to function in the presence of the biomass and to show purity levels equivalent to an isolated active substance. The biomass itself needed to be shown to be free from all living organisms while also having low levels of impurities or process termination substances.
Mrs Kirby-Smith showed that for recent applications of feed grade antibiotics, over 70 per cent of regulatory questions were related to manufacture, in-process controls or specifications. She also reminded applicants of the need for quality and clarity in their submissions, especially in taking responsibility to ensure that expert reports and drug master files conformed to the standards in the rest of the documentation.
In the future, producers of active substances for pharmaceuticals would be subject to good manufacturing practice (GMP) inspections. The guidelines for feed additives controlled under Directive 70/524/EEC were being updated (amending Directive 87/153/EEC), and manufacturers would be subject to inspection. - Contributed.

Issues for the microbiologist

Challenges for the microbiologist were addressed by Dr EDMUND LEE (Medeva). He suggested that analytical microbiologists needed to be closely involved with the process and to understand the main driving forces of business, regulatory affairs and good manufacturing practice so that they could both actively work to control and evaluate levels of contamination and also advise on process design, raw materials sourcing and testing and facility design.
Key to their role were the issues of contamination and regulatory compliance, where the identification of their potential sources and prevention were as important as the actual testing and monitoring. Dr Lee's view was that contamination needed to be controlled right from the start, from screening, quarantine, microbiological and genetic stability testing of cell lines to in-process and end-product testing. Every effort was required to prevent cross-contamination of cells/products and to reduce process and environmental bioburden through process design and continuous improvement of quality and logistics-related issues. Eliminating the use of antibiotics was also seen as a contamination prevention measure, since masking of contaminants could be avoided.
Equally important to microbiologists was their role in identifying and developing rapid methods of detecting micro-organisms. Finally, as far as microbiologists were concerned, good microbiological practice attributed to good laboratory practice, good manufacturing practice and good business.

Challenges for the analytical chemist

Mr PETER WHYMENT (Eli Lilly & Co) described issues for the analytical chemist covering fermentations from 5 to 150,000 litres capacity capable of generating hundreds of samples. One challenge was to identify which components of the complex media were actually needed.
Mr Whyment reviewed the analytical tools from the 1970s to the present, ranging from thin layer chromatography through high performance liquid chromatography (HPLC) to on-line near-infrared spectroscopy (NIR) and multi-channel auto analysers for key nutrients. With huge numbers of samples there was an increasing need to automate testing, but this had to be balanced against a real need for the result. Also there was an opportunity to develop existing methods and make them more efficient, often by shortening run times especially for HPLC where shorter and narrower columns might be used.
Yield improvements were the real benefit of a successful process development backed up by a well designed analytical programme, including some novel solutions to sampling on-line. But this could be a significant cost, and the most effective improvements were therefore obtained on those processes that were as yet underdeveloped. As if to exemplify warnings from earlier regulatory speakers, he showed that yield improvements could be accompanied by changes in related substances, which then needed identification and validation of equivalent safety.
Mr Whyment concluded by saying that increased analytical demand with restrictions on headcount meant there was a greater need and reliance on automated techniques, rationalisation of assay procedures, streamlining of testing and moves to implement more on-line testing (eg, by NIR) at the site of production. He also highlighted the benefits of developing academic partnerships where expertise could be shared and allowed to look at novel analytical applications off-line from the pressures of commercial production.

Problems in handling mycotoxins

Problems associated with handling and testing for mycotoxins were presented by Dr HUGH DELARGY (Medicines Control Agency). The concerns arose from contamination of fermenter feedstock materials, eg, aflatoxins in soybean meal, ochratoxins in oats, barley soybeans and wheat, zearalenones in oats, wheat and barley, and patulin in potato dextrose broth. Because of the extreme toxicity of these mycotoxins, a small pocket of high contamination could affect a huge consignment of material. This caused difficulties in sampling and additional problems of airborne contamination if the process called for size reduction.
Handling of positive controls for test procedures also became a major health and safety issue. Extreme care and high containment were needed in preparing samples, as they were highly toxic mutagenic carcinogens. The materials were often photolabile and so ultraviolet methods could not be used (eg, for visualising thin layer chromatography [TLC]). However, the materials were often fluorescent or could be made to fluoresce. Current official methods of AOAC International [formerly the Association of Official Analytical Chemists] were TLC, liquid chromatography (fluorescence) with enzyme-linked immunosorbent assay (ELISA), gas chromatography and bioassay. Method development continued, with LC-MS (liquid chromatography with mass spectrometry) being useful in both limiting sample size and being able to test for most mycotoxins in one run, and capillary zone electrophoresis being evaluated for its sensitivity.
However, new compounds continued to present a threat and as recently as 1998 a new mycotoxin, fumonisin, had been discovered in maize which could cause oesophageal cancer in man. Drug substances at risk had to have a "non-detectable" specification at the limit of detection of a state of the art method. There also had to be a limit on feedstocks, which should be routinely tested. [Official methods are to be found as: ISO 6651 and 6870; AOAC/IUPAC 1995 (revision 1997) Chapter 49.]