All opinions are my own and do not necessarily reflect those of Novo Nordisk.
h/t to @Frank_S_David, @scientre, and the LinkedIn Group Big Ideas in Pharma Innovation and R&D Productivity for links and ideas
Biopharma may be ripe for disruptive innovation to come in and overturn their markets but that doesn’t mean it will happen. There are constraints beyond those of pure business, including the simple fact that treating diseases is really difficult and we don’t know as much as we would like about how biology really works. I see today’s biopharma market as a victim of its own success. The 80s and 90s saw the creation of truly life-changing, effective drugs like statins, and that has set the bar high enough that I think we’ve passed the inflection point at which approaches like high-throughput screening are becoming less likely to yield a substantial improvement in effectiveness. I’ve used the analogy before of drug development occurring on an adaptive landscape (Figure 2), with every improvement moving up a hill towards the theoretical perfect medicine at the apex. The higher up the hill one gets, the harder it is to move uphill and most efforts move sideways or down, simply because there’s more territory in those directions. This is a constraint that a disruptive innovation would have to overcome in some way.
Figure 2: The adaptive landscape for drug development. Yes I drew this myself. I would plug the drawing program, except I think they’d probably prefer not to be associated with this image.
In addition the long, long lead time that drug development currently requires creates one of the bigger problems for the industry. Compare the ten to fifteen year turnaround for a drug to the shorter development cycles for software, movies or even airplanes. It’s clear biopharma is on one end of the bell curve of product cycle times, and that means there are too many things that can happen in science, the practice of medicine, the regulatory environment, culture and nature to allow very accurate predictions of markets and results. It’s an industry where the likelihood of black swans appearing is greater simply because development takes so long. With such a lengthy development time, the chance of a disruptive innovation being able to overtake conventional drug development methods is decreased. Unless, of course, the disruptive innovation specifically addresses development time.
Another barrier, at least for development from within biopharma, I’ve already touched on in part 1: how most of the time disruptive technologies do not have a clear market. They simply don’t work or scale, initially, into the market they will eventually supplant. The Innovator’s Dilemma has many good examples, so let me just pick ink-jet printers because I’m about to run out of ink in mine.
Ink-jet printers when they first came out weren’t (and still aren’t) as fast as laser printers. The printing isn’t as crisp. But the footprint is typically smaller and the cost (initially) much lower. Except for stupid ink cartridges. Ink-jet printers turned out to have a different, down-scale market from laser printers, and have done well enough that, at the time of the publication of The Innovators Dilemma, they were taking over some of the markets previously held by laser printers. In the present I think it’s safe to say that, while not supplanting laser printers for all uses, ink-jets have developed a stable market in home printer use.
A last point with respect to whether drug development can be subjected to disruptive innovation is that its market may not let that happen. Because human health is so highly regulated, and given my premise that the true market for biopharma is the regulatory/payer environment, there may not be a way for a new technology to ever get enough traction to meet the minimal market demands. We see this already even with incremental innovation, in which new biotechs often have to partner with a large pharma in order to get through the regulatory process–they have a very difficult time doing it on their own with their limited capital. I’m not sure this will actually be a problem, as I’ll describe below, but it’s worth keeping in mind.
But if we start with the premise that disruption is possible, what form will that take? I think everyone, from investors to patients to CEOs would like to know. My short answer is: I have no idea. None! Well, a few, but I’m pretty sure I’m wrong. Not that that will stop me from speculating as we move forward in part 3. One of the reasons innovation is hard to do is that often what was innovative can only be understood in hindsight. Kind of like in James Burke’s Connections series. Who knew coal tar could be so important? So let me approach this question from a different angle, once again borrowing from the framework of the Innovator’s Dilemma. In general, one of the distinguishing characteristics of a disruptive technology is that it initially meets the needs of a different market than that of the established technology which will eventually be displaced. The new market is one that has a different Value Network than that of the established technology, which allows the new technology to make sufficient money to survive and continue its own process of incremental improvements until it reaches a point of being able to compete with the established technology in its own market.
By the way, in doing this post I’m making the assumption that many people would like to know where the next disruptive innovation is coming from and think about how it might be found. If you don’t, why are you reading this? It can’t be for the artwork.
In any case, maybe the way to look for disruptive innovation is by looking at what markets exist for products or approaches that could eventually do what drugs do. Which, as I alluded to earlier, can be thought of more broadly than simply medicines for diseases. Rather, if one thinks about approaches that can lead to changes in complex systems, the possible markets become much broader–some medical, some not.
So what might the markets be? I’ll just throw out a few: Animal and/or agricultural markets, neglected tropical and third world diseases, patients’ organizations, and industrial bioprocessing.
Starting with animal health and agricultural markets, when it comes down to it there isn’t really that much difference between a cow, a dog, a maize plant and a human being. Laura Strong had a nice post about this on her blog, The Next Element. Animals and plants are all biological systems for which there is a great need to figure out how to change that system to better meet human needs. The needs might be for a heftier cow, a medication to reduce a canine tumor, or an increase in crop production. The Value Network is different because the regulatory environment is not the same, the barriers are lower, and the market broader. Anything that is meant to medically improve human health has to go through rigorous clinical trials to be so labelled. For beef cattle, not so much. The Value Network for animal and agricultural products is focused more on concerns like environmental safety (ie, reduced pesticide and herbicide use), increased efficiency (growth with fewer vitamins, antibiotics, or fertilizer), and public image.
However, biopharma is generally not interested in animal health and agricultural products because the profit margins are much lower and the biology is different enough from human biology that working in this area requires a separate research program and business unit. In fact, we see today that major biopharma are divesting themselves of their pet and animal medicine units.
But there is a need, and technologies will continue to be developed to meet that need. This provides one potential market to support a disruptive way of altering organismal phenotype, as well as a market with a much lower regulatory hurdle. If that technology gets powerful enough, it’s not too much of a leap to suggest that technology could eventually be used in human drug development.
Another potential market that could support a disruptive technology is Global Health. Organizations such as the Gates Foundation and PATH are working toward developing medicines and vaccines to help the developing world, but biopharma, generally speaking, is not. While some pharma such as GSK have made strong goodwill gestures such as opening their compound libraries for screening for diseases of the developing world, and others such as Sanofi-Aventis are partnering with the Gates Foundation for vaccine development, the vast majority of resources in Biopharma is directed toward first world diseases. Again, this arises from simple math: large corporations need large profits to survive. The developing world does not provide the kinds of profit margins that would sustain a large biopharma’s efforts.
The Value Network in the developing world is very different. There the cost factor for medicines is extremely important. Safety is still a major concern, but the burden of disease is also great, which adds weight to finding solutions that are simple to apply. In addition, interventions need to be low in technological barriers, so that they are, for example, highly stable and easy to administer. The regulatory environment may not be as stringent because the emphasis is more on getting medications out quickly to serve a highly unmet medical need. This provides an opportunity for new technologies that are able to find and develop medications quickly, and possibly different kinds of medications or other interventions that will help modify a phenotype.
A third potential market is patients’ groups. One of the effects of the interconnectedness that the internet provides is that patients can now find one another, whether they belong to large or small (or tiny) disease populations. They can build communities. While there has been a large focus recently on treating orphan diseases and many of the biopharma are pursuing cures for patient groups with low numbers of patients, there are still many underserved and unserved groups in this category, including diseases that have such a small number of affected individuals that they might never provide a suitable business case for large biopharma drug development. The revolution in genome sequencing is fueling the discovery of causal variants underlying many syndromes, but translating that to a cure is still far away for most.
The Value Network for these kinds of extremely rare diseases tilts sharply towards speed, efficacy and cost. Indeed, it’s not out of the question to think of motivated patients’ groups taking drug development at least partly into their own hands as the Cystic Fibrosis Foundation has done with Kalydeco in partnership with Vertex. There is danger living on this edge, with the possibility of groups being exploited by unscrupulous types promising cures. However, with organization, self-measurement and reporting via Quantified Self-type tools, and some funding, these groups could support new approaches to drug discovery that try to get to a cure more quickly, albeit with (much) more risk.
One of the interesting ways to think of this is to compare patients’ groups to other peer-to-peer networks that have been showing how connectivity can circumvent traditional business models. Just like AirBnB cuts out the hotels, and Lyft is working to challenge normal taxi regulatory processes, a patients’ group could form its own independent drug development organization to meet its own, very specific need. It could even influence the regulatory environment—for example, as is being done with fecal transplants.
The final market I’ve thought about is outside the realm of medicine entirely and instead covers the field of industrial bioproduction–producing industrial materials through the use of biological actors such as enzymes or cells. Here there may not be as strong an unmet need–the industry is thriving–but I think it’s worth mentioning because it also deals with trying to create value out of inherently complex systems such as bioreactors full of cells in media. The Value Network emphasizes throughput, environmental safety, stability and efficiency. This creates a market for testing different methods for predicting the outcome of perturbing a complex system. And that’s what medicines are also trying to do. A monoculture of cells growing in a bioreactor is a lot less complex than a human body, but that’s the point. It’s a simpler system where innovations for prediction can be tested and refined, providing an opportunity for innovations to mature.
Two other things that are attractive as industrial bioproduction as a market for innovation: the timescale for development could presumably be both rapid and amenable to small scale piloting, which would further support the proof of concept testing of novel ideas. And, many of the biopharma have conceptually similar capabilities in house for production of antibodies and vaccines, meaning transfer of a new technologies into biopharma might be easy via this route.
So to conclude this part, these are just a few of the potential markets in which disruptive technologies might find a home before moving on to disrupt conventional drug development. The main characteristic uniting them is the meta-trait of requiring something orthogonal to the Value Network biopharma currently faces.
Coming up in Part III: What would disruptive technologies look like? We could use some crystal balls.