Precision Medicine; Imprecise Manufacturing and Supply

Many drugs fail to have any beneficial effect in a significant proportion – in some cases more than 50 percent – of the people who take them. It’s also common knowledge that many candidate drugs fail late in their clinical development, after eating up huge amounts of time and money, because they may trigger severe adverse reactions in a proportion of patients. Adverse effects in even a very small number of patients can be sufficient to cause product withdrawal, regardless of the benefits the product offers to the majority. It is now recognized that these systemic inefficiencies, and their significant economic and health consequences, are primarily due to genetic variations between patients. And this recognition has led to the development of tools and techniques to adjust treatments according to genetic background and other individual charateristics – so-called personalized medicine, stratified medicine and similar concepts.

The underlying intent of all of these is to diagnose a condition quickly and unambiguously, and then to match the diagnosis to a specific medicine or therapy, determined in part by the patient’s particular genotype or phenotype. The preferred term for this paradigm is ‘precision medicine’, and it is being intensively pursued, at both the preclinical and clinical level. There is a belief that ineffective treatment, waste and late-stage pipeline attrition can all be addressed if precision medicine approaches are brought to the fore.

However, amongst all the fervor and investment that precision medicine has attracted, some key issues seem to have largely been ignored. In particular, there has been little discussion of the consequences that the introduction of precision medicine techniques will have for medicine manufacture and downstream supply chains. Among the few of us who have considered this, the perhaps unwelcome view is that, within a decade or so, most of the existing medicine manufacturing equipment and facilities – and the downstream supply chain to the patient – will no longer be fit for purpose and will require a radical overhaul to meet the needs of precision medicine.

I believe that the medicines industry is currently not well-placed to address the manufacturing and supply challenges that are heading towards it. Few countries are rising to the challenge of precision medicine, but on a positive note there have been some interesting projects in my home country, the UK. I’m a firm supporter of the UK economy and if the country can get its act together then it has the potential to retain and attract medicine manufacturing in this novel arena, and to become a global leader in the development and application of related, novel technologies. In contrast, the continued support of conventional medicine manufacturing in the UK with little more than incremental improvements and fiscal incentives doesn’t seem to be much of a future-proof strategy. I’m hoping that the UK will grab the opportunity to get good at the new stuff, and start using it.

Uncomfortable truths

First, let’s start with some of my background. I’ve been involved in advanced manufacturing systems development for 40 years, mainly in the application of robotics and related computer-controlled, adaptive processes. I’ve worked across a wide variety of industries and countries, but for the last 25 years I’ve specifically focused on innovation in pharma and biotech manufacturing.

I’ve found that biopharma manufacturing can greatly benefit from application experience in other industries, and that this experience can challenge established wisdom and lead to innovative process methods. Indeed, this cross-fertilization has resulted in the development of many new pharma manufacturing methods, such as the use of robotics to scale up the production of biologics, which have widely been adopted (mainly by US companies).

When I lecture manufacturing engineering undergraduates, I often ask them to consider and contrast the challenges I (and my colleagues) faced when developing machines and processes to make 1000 Cadbury’s Flake chocolate bars a minute (it was really, really difficult) as compared with automating erythropoietin manufacture (piece of cake). The sophistication of the product does not necessarily imply a corresponding level of manufacturing technology challenge; in fact, the inverse can often be the case. I have a strongly held view that medicine manufacturers need to get out more, and be less dismissive of what they might consider to be inferior or non-relevant manufacturing industries.

After working across many industries, I’ve become increasingly aware of the critical importance of defining and adopting a clear and relevant manufacturing strategy. Essentially, you must define how manufacturing should best support – at a strategic level – the overall business objectives of a company. You must ask, what business-critical capabilities do you need in your manufacturing section, and what are the implications of delivering those capabilities? This kind of analysis requires attention to issues that are much more pervasive and long-term than just a continual pressure to reduce costs. An example of where manufacturing capability really does matter at the strategic level is commercial jet engine production. Rolls Royce has made very significant investments in people and facilities, so as to support a strategy of continual advanced process development that links intimately with, and enables, new product development. In particular, Rolls Royce’s ability to manufacture advanced turbine blades is a critical strategic factor; the blades are made of a sophisticated alloy formed as a single crystal in a precise, complex shape, which can then operate continually, at very high stresses, in gases above the melting point of the material. Essentially, their manufacturing strategy means they can make things which few, if any, other companies could ever do.

So an uncomfortable question for the medicine manufacturer is, what manufacturing capability do you have that is unique? What do you make that couldn’t readily be made by others? An insightful, if cynical, industry senior once told me that current pharma manufacturing strategy could be summarized as “never run out” – the implication being that intensive use of capital resources and rapid responsiveness were not particularly necessary or business-critical, provided there was always plenty of stock in the warehouse and downstream. Hence, a tablet typically may be one to two years old at its point of use, and some of its component materials may be two years older than that. It’s hardly a paragon of lean, responsive manufacturing!

A second aspect of medicines supply that seems to be way behind the best practice curve is the process by which drugs currently reach the patient. For example, as I once again hit the “purchase” button when shopping online with Amazon, I have every expectation that I will receive my goods correctly, within 24 hours, and at apparently little or no cost for the logistics by which they reach my door. Yet my repeat prescriptions (which are also initiated by an Internet request) can take up to three days to arrive at my surgery’s pharmacy. And I still have to go and collect them...

I fail to see what the drug supply chain provides that an Amazon-style approach couldn’t do much better; service level data would probably show that errors are much less frequent in the Amazon model than in supplies distributed via pharmacies. In the US, pharmacy benefit managers have been supplying drugs via mail order direct to patients, under volume contracts with health insurers and major employers, for many years – and a significant proportion of the total US market uses this form of supply chain, without apparent difficulty. So clearly it can be done in other markets too. For countries like the UK, where there is a high concentration of people on a small island under a national health system, it should be easy to achieve. Technically, a single national mail order pharmacy would be perfectly feasible (although I think that some influential – and unnamed – stakeholders in the current system might resist this kind of development for their own business reasons).

To summarize so far, there are two specific issues that I am identifying in the pharma industry; the senescence of process technology in pharmaceutical manufacturing, and the tardiness and inflexibility of the supply chain to the patient. These two issues are important to focus on because they have the potential to adversely impact the delivery of precision medicine. Precision medicine will be ready to move into broad practice within the next decade, and a failure of the manufacturing and supply chain to innovate in readiness to meet that challenge will be damaging, both in economic terms and in delivering the most effective and affordable outcomes.

No doubt manufacturers will be reluctant to change their traditional ways, but change is a great opportunity. Without precision medicine and the anticipated changes to medical prescribing, the manufacturing and supply chain status quo also would have little incentive to transform; in fact, there would be continuing comfort in sticking with the devil you know. After all, it’s been good enough (for some people) for the last 60 years.

But isn’t it time to bring medicine manufacturing out of the dark ages? If we appropriately prepare for the forthcoming changes then we’ll be able to reach a world where manufacturing is as strategically important (and exciting) as it is for Rolls Royce. A famous saying from 1962 about a big challenge was, “We choose to do this, and the other things, not because they are easy but because they are hard”. And I think that it’s time to start doing some hard things in medicines manufacturing.

Striving for the future

So what does the adoption of precision medicine mean for the medicine manufacturer of the future? I think there are a number of important points to make.

One. There will be a significant increase in the number of products to be made and distributed. Some would like to think better diagnostics and patient-specific data would just mean better matching of patients to existing products, but my discussions with drug researchers suggests that this is too simplistic a view. You could imagine a scenario in which a pharma company developed a family of novel compounds, perhaps addressing a novel mechanism of action for a given indication, where each family member targeted specific, differing genotypic traits, but was associated with serious adverse reactions in patients with inappropriate genotypes. It’s difficult to accurately assess right now, but an order of magnitude increase in product diversity is not inconceivable. It’s interesting to imagine what a conventional retail pharmacy might look like offering that many more products on demand, and on an absolutely error-free basis. I suspect that retail pharmacy and precision medicine may prove to be incompatible partners, in which case supply of medicines direct from the factory or warehouse to the patient might prove to be the only practical approach to handling the increased product portfolio.

Two. Increased product diversity will, of course, significantly impact primary and secondary processes at the drug manufacturer. Batch sizes will shrink, requiring the development of more agile systems, capable perhaps of making several different products on the same day, while still maintaining absolute control and integrity of each different product and process.

Three. There is a clear need for innovation – but who is going to conceive and create the process technologies and equipment that can deliver these new levels of performance and flexibility in the coming decades? Current pharma manufacturing technology mostly comes from well-established, specialist companies. The main players offer long-established methods and designs, and there has been little incentive or customer demand for them to embark on radical innovations.

Anyone who studies disruptive technology theory (see Clayton Christensen’s writings, for example) will tell you that radical innovations almost always come from unexpected places, typically players with different skill- and mind-sets that are unconstrained by established wisdom and convention – the light bulb wasn’t the result of increased R&D on the candle. Therefore, your existing supplier base may not always give you the technology you need for the future. It also means that there is an opportunity to create new manufacturing technologies and to nurture the companies that commercialize them, such that they leapfrog the competition. I believe the underlying core technology capability exists, but we might have to look in unfamiliar places to access and exploit it. And successful exploitation may require more cross-functional collaboration than we are used to; in particular, bioscience and engineering may have to operate on more equal footing in the pharma company of the future.

Five. In order to meet these future objectives, pharmaceutical manufacturing will need to diversify its skill set and attract high performers in new areas of expertise. That is, people who today might not associate medicines manufacturing with the challenges, pace and excitement found in other manufacturing industries, such as electronics or aerospace… or even vacuum cleaners. The pharma manufacturing industry has an image problem, and that will have to be addressed if we don’t want the capability to innovate to be constrained by an inability to attract and excite the best talent.

And six. What I am describing as the future conflicts with current regulatory guidance – it wouldn’t be innovative if it didn’t. Regulators get a lot of bad press in our industry and are often being cited as a barrier to change, but I suspect that in many cases the regulations become an easy reason to maintain the status quo – and a quiet life for the manufacturer. The reality is that the onus is on the developer/user to demonstrate that the challenge to conventional practice they are proposing offers important patient benefits, while also providing enhanced safety and quality. There’s a raft of powerful sensing and control technologies out there, which can help you to achieve this, if you only know where to look.

The underlying point across all of the above is that the delivery of precision medicine involves manufacturing challenges that are not simple, but which are certainly tractable, given sufficient commitment and resources (and an early enough start). It does mean, however, that you must begin with a clean sheet of paper rather simply looking for incremental enhancements to current methods and equipment. We are at the top of the “S” curve with current approaches; in other words, further investment will only yield modest gains. It’s time to jump to the bottom of the next S curve to get step-change benefits from a new way of doing things.

As I’ve mentioned previously, the UK is involved in a lot of activity to promote medicine manufacturing – and I’m pleased to see that the UK government recognizes its retention and growth as a national industrial priority. My principal hope though is that the emphasis of this support is targeted at delivering the manufacturing and supply needed for our future healthcare, and is given the leadership, funding, time and commitment to achieve that outcome.

Escaping the box

Finally, let me share a piece of personal history, and perhaps an indicator of what might be possible if you can fight your way out of the box. About 20 years ago, my company of the time became, in effect, the in-house advanced manufacturing automation arm of Merck & Co in the US. This was a strategic move instigated by their senior executives, who had identified that their existing equipment vendor base did not have the new skills and resources needed to create the innovative processes the future might call for. We undertook a diverse range of exploratory projects, culminating in the concept of a production line that could simultaneously pack many different tablet types into bottles, with each bottle being specific to a patient prescription, having a variable count and a unique label and leaflet, which were printed on-line. Every tablet would be checked for size, shape and color, and subsequently active content and dosage level too. Every prescription also had its own bar code and batch record. The tablet count was verified by two independent methods, with no product-specific tooling. Different bottle sizes could be handled simultaneously, and at the end of the packing line, prescriptions could be collated and placed in envelopes ready for collection by a carrier to enable delivery the next day. Not unreasonably, there was some skepticism about the concept both technically and commercially. So, with the direct support of Merck’s CEO, we built a demonstrator line that produced 150 custom prescriptions per minute and delivered all the functional capabilities outlined above. Sadly, due to a complex mix of issues, it progressed no further. But if this intrigues you, take a look at the system on YouTube (http://youtu.be/la6c6qOW4zE).

Given that we developed that system with 1995 technology, imagine what we could do today.

Richard Archer is the managing director of Two BC Ltd, a non-executive director of the UK Biocentre, chairman of the Engineering and Physical Sciences Research Council Centre for Innovative Manufacturing in Regenerative Medicine, and a senior industrial fellow at the Institute for Manufacturing at the University of Cambridge, UK.