As the market moves towards more targeted and niche products, Paul Titley, ceo of R5 Pharmaceuticals, looks at the advantages and challenges of producing new drug candidates on a very small scale
The pharmaceutical industry rightly spends a lot of time and money on making sure scale-up and technology transfer is efficiently organised, and recorded. Technology transfer is always about moving practical endeavour from one site to another, and often associated with a change of scale, usually an increase. But at the other end of the spectrum - the tiny batch end - very little seems to be written or practiced. When formulations and processes are used to make the first medicines from discovered molecules, what can be done to make small but useful batches?
The usual problem with tiny batches is that there is no obvious equipment to use. The commercial representatives on the project team may also query how the process used at the tiny scale relates to the large and huge scale. The most honest reply is to say it doesn't relate at all. It may also be the opportunity to challenge the "lab version" models of many items of equipment. It is very difficult to reduce the features of commercial pieces of kit to something you can use to obtain accurate scaleable process data. Exceptions may include drying or coating temperatures. It is unlikely that mixing times, paddle speeds, drying times, etc will be directly scaleable from your lab version equipment.
Twenty-five years ago at Interphex 1982, Richardson (Boots) presented a paper on sweep volumes1. He discussed various ratios measured from a range of mixer-granulators (25, 100, 400 and 600 litre models). To effectively build the "same" machine in different sizes what properties or dimensions are scaled? It could be the paddle tip velocity, the paddle speed, the chopper speed or size, or the volume swept by the locus of the paddle. Richardson argued that the ratio of the volume swept per second compared to the volume of the mixer was a useful indicator of scalability. This means it should remain the same from one model to the next for processes to be easily transferred from one model to the next. The ratios were 2.74:1.69:1.14:1.00 for the 25l to 600l models. His paper offered this data as one reason why processes developed on one model don't transfer perfectly to another model. The same examination is probably feasible on any range of equipment. My purpose in including this is to suggest that it is safe to abandon processing concerns at the tiny scale. Even when your process is stabilised at the medium scale there will still be plenty of work to be done in transferring it to the commercial scale process.
Early pharmaceutical development work is focussed on producing a medicine from a molecule - something tangible that a patient can take. It is also the first time that the "owner" of the molecule gets to see what the final product might look like. Not many capsules, tablets, vials etc are required to achieve this objective, so why make large overages. The desire for tiny batches may also be driven by the lack of compound. Making tiny batches will enable you to "peek into the future" and give you the opportunity to examine some "what-if" scenarios. Such manufacture is not an exact science and care must be taken not to treat the resulting data as perfect. Such batches are not going to be able to be described as "made to GMP"; this is "handicraft pharmacy".
The easiest tiny batch is the capsule. We can all fill a single capsule. Capsugel supplies the Xcelodose Precision Powder Micro-filling System - filling equipment with startling precision to place tiny amounts of drug into single capsules. The result is a product that can be used in the clinic, bypassing process development, but also bypassing formulation development.
Formulated powders can be mixed in small blenders or tiny ball mills and then filled into capsules using a steady hand or by using manual filling units.
However the most common pharmaceutical form is the tablet and they are not so easy to make in ones and twos. An instrumented single punch press can provide both a tiny batch, but also data to hint that your formulation is at least on the right lines. Nevertheless small biotech companies that have Phase 2 success as their objective are unlikely to share your enthusiasm for this type of experimentation. The granule or powder to make the tiny batch of tablets can be mixed in small food mixers or in small blenders, such as the Turbula. Fluid bed drying is possible with care, but this introduces another aspect that you may want to avoid - cleaning validation.
When keeping the costs of manufacture to a minimum it is perverse to add to these with analytical chemistry costs to establish that your equipment is clean after use. Food mixers are so cheap that they can be discarded after use, but expensive pharmaceutical equipment can rarely be casually discarded!
Drying of wet granule can be done on trays, although this will often produce a harder granule than that from a fluidised bed. It will almost certainly require more sieving/grinding to get the granule to suitable size range for you to compress it. Tablets can be made using hand operated machines or simply in an engineers vice.
Data derived from such tiny batches will give you a guide to the future. It is possible to manufacture melt granulations in traditional pestle and mortars, dry in air, manually mill and sieve, then compress using a vice and still get useful dissolution data to demonstrate the principle of the formulation approach.
Liquid fill capsules in hard shells can be made in singletons and sealed using gelatin or HPMC solutions applied with a small brush. Softgels stand out as "not possible to make tiny batches".
Many tablets and some capsules need to be coated for a variety of reasons: to control the release of the active, to mask the taste, or to protect those handling the product or a combination of reasons. Very small coating pans do exist (e.g. LDCS from Vector Freund) however there are cost implications as mentioned previously. Tablets can be hand painted with coating solutions, but this is not as easy as it sounds. Regular film coating is sprayed on and is built up gradually from a series of minuscule splashes; the resulting coat is smooth and thin. Hand painting will produce a shell-like covering with the risk that it is too thick and each side has to be painted separately to allow the other to dry. However with care some satisfactory products can be made this way.
Liquid products are easily made at the tiny scale, and these include injectables. It is possible to autoclave a single vial or ampoule (to assess decomposition). It is also possible to lyophilise such tiny batches.
In conclusion the tiny batch offers a means of examining what-could-be and possibly what-might-be. Such manufacture requires care and experience, it does not replace what must come later, but it does allow you to examine where you should spend your valuable development budget.