Traditional small molecule pharmaceuticals are synthesized via highly reproducible chemical processes. The resultant compound is patented based on its atomic structure rather than its manufacturing process because of the fidelity of these reactions and also because determination of the purity and composition of such small molecules is routine. In contrast, protein therapeutics are on average 100-1000 times larger than small molecules and logarithmically more complex. They are produced in recombinant organisms, usually bacterial, yeast, or mammalian cells, and the protein is purified to homogeneity via biochemical methods. These methods are often lengthy protocols of which any or all of the process can be proprietary. In addition, the utilization of living organisms as miniature factories results in an inevitable heterogeneity in the manufactured therapeutic. The final product can be effected by minute alterations in protocol, and stringent attention must be paid to the necessities and behavior of the co-opted organism.
The bottom line is that not only are protein therapeutics themselves overwhelmingly more complicated than traditional pharmaceuticals, the degree of randomness and consequent difficulty involved in their production is many orders of magnitude higher than the analogous small-molecule synthetic chemistry manufacturing methods.
Given the complexity and heterogeneity described above, its no surprise that in the early years of biologics the philosophy that "process defines product" governed regulatory actions. The result was that the FDA required a single biotechnology company to to obtain a Product License Application [PLA], and an Establishment License Application [ELA], in addition to performing pivotal phase 3 trials using the same facility used for final commercial production. Fortunately for the Biotech industry, this was replaced with a single Biologics License Application [BLA] through the FDA Modernization Act [FDAMA] in 1997, and companies were also permitted to change, or more importantly, outsource their manufacturing process so long as the resultant therapeutic was shown to be comparable.
Early biologics such as Amgen's (NASDAQ:AMGN) recombinant erythropoiten and Genetech's human growth hormone suffered from a supply deficit even as they proved to be commercial successes. However, it was Immunex's (acquired by Amgen) Enbrel, released in 1998 as a treatment for rheumatoid arthritis, that would become the prime example of a biologics manufacturing shortage. Supply rapidly outstripped demand leading to patient waiting lists and shortfalls that continued until the end of 2002. This event opened the doors for competing biologics like Centocor's (acquired by J&J) Remicade, which did have enough manufacturing capacity, and may have ultimately may have cost Enbrel's makers more than $200 million in lost revenue.
These above events gave birth to a boom in the industry of Biologic Contract Manufacturing Organiztions [CMOs] almost overnight. Building a new biopharmaceutical plant costs hundreds of millions of dollars and can take up to 5 years, many biotechnology companies are reluctant to make that kind of investment to support a therapeutic candidate still in clinical trials. Business vacuums don't exist for long, and in only a couple of years the concern over biopharmaceutical manufacturing capacity had died down considerably. However, CMOs are here to stay and their growth outlook is generally accepted as favorable. Big pharma and larger biotechnology companies are building their own facilities for biologic manufacture, or even adopting a shared-capacity strategy. Meanwhile, smaller biotechnology firms, especially those with only a few candidate therapeutics will continue to rely on CMO's facilities and expertise as they focus resources on product development and establish a proof of concept in clinical trials.
Just as CMOs are considered by the biotechnology industry as being a bright side of biologic manufacturing, their close relative, generic biologics manufacturers represent the dark side. These off-patent versions of protein therapeutics, dubbed biogenerics, biosimilars or follow-on-biologics are already a reality in Europe and elsewhere, and they are on the horizon in the United States. This is the current hot topic of debate surrounding biologics, and although arguments can be made on either side as to how long it will take for either the US legislation to pass or for the developing nations to bring their cGMP facilities up to speed, most will agree that biogenerics and outsourcing are only a matter of time.
With respect to the biogenerics market, I consider the business of evaluating either domestic legislative or foreign compliance risks particularly volatile. The above retrospective does however, firmly illustrate the overwhelming market pressures in the biotechnology industry as a whole to not only discover new protein therapeutics and biologics, but also to produce them on a industrial scale in an increasingly more cost-effictive manner. This thesis is restated in a recent article written by the senior VP of technical operations at Wyeth.
To get a glimpse of what the future may hold for the biopharmaceutical industry, one need only look back to the transformation that took place in semiconductor manufacturing. Similar to the biotechnology industry, the technology for semiconductor manufacturing was initially highly specialized and expensive. Competitive pressures and the need for large-scale production required the construction of large plants, at costs that were prohibitive for most industry companies. The investment in such large plants led to a compromise in the ability to rapidly respond to new technological advances. To better respond to markets and compete with lower cost operations in Asia, semiconductor companies began to form consortia to share capacity and hire contract manufacturers. As in the biotechnology industry today, shared capacity in semiconductor manufacturing was only possible through the standardization of processes and technology. Technology standardization became more firmly established as the small number of companies, which held the dominant intellectual property required for the design and manufacture of state-of-the-art semiconductors, became the industry leaders.
Although there are many obvious differences in producing protein therapeutics versus microprocessors, the most notable is that protein production is not easily standardized. In this regard each protein product will have a customized process to some degree.Yet I believe the comparison to be apt in terms of market opportunity. Just as Intel benefits whether Microsoft or Google become popular, developers of technologies which can successfully and generally reduce the cost of protein production should stand to gain handsomely regardless of any of the above market uncertainties.
The bottom line is that those technologies which can facilitate the production of biologics more economically will equate to a competitive edge in an industry beset on all sides by an imperative to reduce manufacturing costs.
In this manner I also hope to be able to predict the long term valuations of companies supporting the biomanufacturing industry. Applied to medium and large cap stocks, I would expect Invitrogen (IVGN) to post better than expected earnings tomorrow. In a likewise manner I expect Thermo Fisher (NYSE:TMO), GE Healthcare (NYSE:GE) and BD biosciences (NYSE:BDX) to continue to outperform in their life sciences departments into the forseeable future. Pall (NYSE:PLL) and Millipore corporations (MIL) have had nasty tax problems and poor Q2 performance respectively, but the above thesis predicts that their product technologies will continue to show strong value and also present a long term investment opportunity. Internationally, Cobra Biomanufacturing [LSE:CBF.L] and Sartorius AG [XETRA:SRT.DE] among others, meet the criteria outlined above.
Disclosure: I am long shares of TMO