During the past two decades the biopharmaceutical industry has been facing an innovation deficit, characterized by increasing research & development costs and stagnant productivity. From its inception, biotechnology has been expected to counter this deficit by its revolutionary science-based approach to drug discovery. For this study we gathered patent and product data related to the technological development of the first two biotechnologies: recombinant DNA technology and monoclonal antibody technology. We studied the technological lifecycles of these technologies in terms of scientific discoveries and inventions as well as product innovations. Results indicate that over the years inventions related to these technologies have simultaneously become less radical and less valuable. Furthermore, our analysis shows that these biotechnologies have reached a stage of technological limit or saturation, which may be followed by an innovation cliff. Now, more than ever, it is crucial to examine new strategies and opportunities for value creation, capturing, and delivery, within the biopharmaceutical industry.
The revolutionary characteristic of biotechnology is the fact that it is derived from advances in fundamental science, and can be used for discovery and development of new products to fulfil unmet medical needs. The rise of biotechnology transformed drug discovery and development from traditional pharmaceutical target screening to a science-deductive process [1,2]. Consequently, it became possible to target new leads based on the understanding of complex biological systems.
From the first technological breakthroughs in the 1970s, high expectations arose that biotechnology would radically improve drug development and generate new classes of biological products. Moreover, specific biotechnological products were expected to counter declining pharmaceutical productivity [3-5].
Contrary to those initial expectations, several researchers have since suggested that those optimistic expectations of biotechnology are unsupported by empirical evidence [
Even with a dramatic five fold increase in research & development (R&D) spending there appears to be no effect on New Chemical Entity (NCE i.e. New Molecular Entity) production, resulting in a pharmaceutical “productivity gap” (
Evidently, pharmaceutical firms are in need of innovation to increase productivity. Therefore, it is important to study the innovation patterns and lifecycles of individual biotechnologies. Such specific patterns can be examined using technology forecasting, a useful tool for identifying phases of a given technology’s lifecycle [14,15].
In this paper, we examine the patterns of innovation regarding the first two major medical applications of biotechnology: recombinant DNA (rDNA) technology and monoclonal Antibody (mAb) technology. These biotechnologies have generated a sufficient number of market-
able biological products that are currently available as prescription drugs. We propose that identifying and analyzing patterns in biotechnological innovation and product development is an important prerequisite for defining optimal innovation strategies needed to improve new product development and value creation in the biopharmaceutical industry.
The application of biotechnologies in medical product development has a relatively short history. The first publications on successful intracellular production of rDNA appeared in 1972 and 1973 [16-18]. In 1974, Stephan Cohen and Herbert Boyer from Stanford University applied for the first patent on rDNA [
In the late 1970s/early 1980s, private companies such as Genentech began to focus on rDNA technology [
The potential for life-saving cancer treatments due to rDNA technology caused a second wave of innovation in biotechnology involving mAb technology [
The two closely related technologies (rDNA and mAb) quickly became efficient methods of producing comercially important substances. Wright (1986) [
According to Garcia & Calantone (2001, p.112) [
The second important aspect of the above described definition is that innovation is an iterative process and therefore includes the introduction of new innovations on the one hand, and the reintroduction of improved innovations on the other. This brings us to the need to classify innovation according to various degrees of innovativeness, distinguishing, in particular, between radical and incremental innovation [27,35,36].
It is generally presumed that a technology follows a certain pattern throughout its lifecycle. The technology saturation-curve (S-curve) method of analysis has been described and employed to retrieve information on the lifecycle phase of a given technology (
According to Ernst (1997) [
Technologies can further be classified according to two dimensions, namely the integration of the technology in products or processes and the competitive impact of the technology [
comes a key technology. Subsequently, over time, the technology starts to lose its degree of competitive advantage and it becomes a base technology. At this point saturation or technological limit is reached.
The technology-forecasting tool is useful because it indicates the current life cycle phase of a technology, allowing companies to strategize for the future [14,15]. For example, when Chen et al. (2011) [
Patent applications are perceived as an important indicator, since patent analysis reveals information on historical developments of the technologies investigated in this study. The patent data for this study was gathered from several patent databases using AcclaimIP patent analysis software1. The acquired data was compared with data directly gathered from the World Intellectual Property Organization (WIPO)2, the United States Patent and Trademark Office (USPTO)3, and Thomson Reuters’s Derwent Innovation Index4. Over the period of 1980 until 2011, we gathered a total of 7350 patents regarding mAb innovations and 9111 patents regarding other rDNA innovations. Patent data is readily available and categorized according to a system of international patent classification (IPC). The IPC is a complex hierarchical classification system encompassing all areas of technology and is currently used by industrial property offices in more than 90 countries [
Recombinant proteins can be divided into various subtypes, with monoclonal antibodies being the largest subtype. Therefore, we decided to divide the sub-types of recombinant proteins so that two different biotech trends could be plotted separately. Falciola (2009) [
In addition, data on biopharmaceutical products was gathered by means of literature research and database development using the Food and Drug Administration (FDA) database5, the FDA Orange Book and the Medtrack database6.
When analyzing patent data, citations form an important source of information. There are two types of citations: backward and forward citations. The former refers to patents that have been cited by the patent in consideration [
(Results shown in Figures 3(a) and (b)).
Where nk is the number of patents in year k and xik are the number of backward citations for patent i in year k. yik are the number of forward citation for patent i in year k and ai is the age of patent i.
The annual accumulation of patents in a specific area of technological innovation yields valuable information regarding technological lifecycle patterns and develop-
ment phases of the respective technology [15,37]. The technology S-curve was constructed by plotting the cumulative number of patents against time according to the file dates of those patents. Similarly to an S-curve based on patents, product introductions related to a specific technology can be plotted cumulatively against time, following a patent S-curve with a time lag of several years due to R&D. Furthermore, cumulative revenues generated by these products help gain insights into current returns on investments in these technologies as well as the future potential earnings from the respective technology related products. Analysis of these three independent parameters resulted in the curves as shown in
Conversely, the average annual trend of forward citations is decreasing in
In
By comparing this technology S-curve to the patent citation trends (
Overall, 81 recombinant protein products have reached the US market. 31 of these concerned monoclonal antibody products.
Similarly, revenues generated by the products included in our analysis have been plotted cumulatively against time (
At a relatively low level of 5% of global pharma sales over the past 17 years, we show that the first biotechnologies have reached a stage of technological limit. New patents related to these technologies are becoming less radical and less valuable and the technology S-curve analysis shows that the technological development currently finds itself in a saturation phase. In addition, the product
curve appears to be reaching a plateau as well, making it difficult to expect future growth in product introductions generated by these technologies. On a positive note, our results indicate that revenues generated by biopharmaceutical products are still growing. However, given these results we conclude that these individual technologies will not live up to the expectation of biotechnology at its inception.
This conclusion gives a quantitative basis for earlier assumptions, which projected that the biotech output during the first years of the 21st century would not be sufficient to make up for NCE deficits [7,9]. According to our results, even less recombinant proteins and monoclonal antibodies eventually reached the market than was projected in these studies [7,9]. In the late 1990s and early 2000s other authors argue that it was too early to tell whether the structural industry changes triggered by biotechnology, would measurably affect industry productivity [7,47]. In hindsight, we can now conclude that so far, biotechnologies exerted little impact on overall pharmaceutical productivity. Moreover, the first biotechnologies that actually generated marketable products are already reaching their technological limits. Subsequent biotechnologies (i.e. combinatorial chemistry, cell-based assays, bioinformatics, genomics, pharmacogenetics and gene therapy) have not yet led to an increase in industry productivity either. In addition, the costs of developing a single innovative compound have risen from 750 million USD between 1995 and 2000 to 1.3 billion USD between 2005 and 2010 [7,26]. The R&D costs for a single marketable product are expected to grow well beyond 2 billion USD, considering current R&D expenditures [48,49].
It seems that the currently employed traditional pharmaceutical blockbuster business model may not be fully applicable to science-based technology and innovation, and the changes it caused in drug discovery [8,50]. Our results imply several scenarios and developments within the industry that may involve different strategies and new opportunities.
ImplicationsThe main conclusion of this paper implies a rather pessimistic scenario for early biotechnologies as the saturation or the maturity phase might be followed by an “innovation cliff” [
Another scenario might involve further development and innovation with respect to initial biotechnologies. During the growth/maturity phases of technological development it is useful to suggest a shift in focus regarding innovation towards the development of new technologies. In reference of the theory regarding technological development and innovation S-curves, rDNA and mAb technologies may have functioned as base technologies [
However, in discussion of the S-curve concept, which is broadly embraced in strategic literature, Sood & Gerard (2005) [
Regardless, it is fairly urgent and important to consider new strategies and opportunities for increased value creation, capturing, and delivery. Extensive exploration of such strategies is beyond the scope of this paper. However, we will briefly discuss the implications of two suggested strategies that might yield significant opportunity for value creation.
According to some, an opportunity resides in reinventing the traditional pharmaceutical business model with respect to diagnostic-drug linked products (i.e. theranostics) and “personalized medicine” [50,53,54]. Newer technologies such as genomics and pharmacogenetics can enrich clinical research by defining patient groups with the most favourable risk-benefit ratio, making it easier to statistically determine efficacy, safety and appropriate dosage of a so-called theranostic in development [53-55]. Thereby, such technologies can function as a “key resource” for a reinvented pharma business model. However, there are two simple but important determinants that form the basis for the current “blockbuster model”; 1) very high and increasing new product development costs, as discussed earlier; and 2) very high attrition rates and thus high risks in new product development. Regardless of the technological possibilities of theranostics and personalized medicine, these determinants remain, and have to be met in any “new” business model.
Perhaps a more realistic opportunity for short-term exploitation comprises industry convergence with the conventional and functional foods sector. Upcoming markets such as the functional food market [56,57] and medical nutrition market [
In conclusion, we show that biotechnological innovation with respect to the first two biotechnologies has saturated. By three independent parameters we have identified the growth, maturity and saturation stages, internally validating the S-curve for these biotechnologies (
We are thankful for support from the Athena Institute at the VU University led by Prof. Dr. J. Bunders. We are grateful for the contribution, through discussions, with Linda van der Burgwal, MSc. and Esther Pronker, Ph.D. In addition, we would like to thank Boudewijn Drost for his contribution regarding data collection and ttopstart B.V. for providing access to certain databases. All authors have declared to have read and approved this final manuscript.