Abstract
Through systematic collection and trending of pharmaceutical data, operational evidence to verify existence of 14 factors affecting the ongoing pharmaceutical transformation has been compiled. These 14 factors are termed transformation triggers. The theoretical evidence in support of these triggers is carried forward from a systematic review of the literature that was conducted previously. Trends in operational evidence and the associated theoretical evidence were compared to identify areas of similarity and contrast. Areas of strong correlation between theoretical evidence and operational evidence included four transformation triggers: a fully integrated pharma network, personalized medicine, translational research, and pervasive computing. Key areas of contrast included three transformation triggers—namely, healthcare management focus, adaptive trials, and regulatory enforcement—for which the operational evidence was stronger than the theoretical evidence.
LAY ABSTRACT: The intent of this paper is to provide proof to demonstrate if there is any operational evidence that supports the 14 transformation triggers previously identified during the theoretical part of this research. The theoretical evidence for these triggers was carried forward to this paper for study from an operational perspective. The practical evidence established in this paper was compared with the corresponding theoretical evidence to identify areas of similarity and difference. This resulted in four triggers that had strong relationship between operational and theoretical evidence; they are a fully integrated pharma network, personalized medicine, translational research, and pervasive computing. The areas of difference included three transformation triggers for which the operational evidence was stronger than the theoretical evidence. These were healthcare management focus, adaptive trials, and regulatory enforcement.
- Pharmaceutical transformation
- Pharmaceutical quality
- Fully integrated pharma network
- Personalized medicine
- Translational research
- Pervasive computing
1. Introduction
This paper is the third in a series that explores ongoing transformation in the pharmaceutical industry (The word pharmaceutical collectively refers to pharmaceutical and biopharmaceutical companies) and its impact on pharmaceutical quality from the perspective of risk identification. The 14 transformation triggers presented in this paper are findings of the systematic review of literature performed by Shafiei et al. (1, 2) that provide theoretical evidence in support of these triggers and ranks their relative importance with respect to pharmaceutical transformation. The review also provides a high level description of operational differences in regulatory approaches between the United States (US) and the European Union (EU). Among the 14 transformation triggers, four triggers—fully integrated pharma network (Trigger 2), personalized medicine (Trigger 3), translational research (Trigger 5), and pervasive computing (Trigger 14)—were found to be the most prevalent within the articles studied. The identification and ranking of the transformation triggers was achieved through a six step process (2) that included selection of primary articles, review of the primary articles, selection of derived articles, testing for article diversity, searching for transformation triggers in all articles, and ranking of transformation triggers. Selected articles were reviewed in detail focusing on views, opinions, actions, and evidence that provided topics on pharmaceutical industry evolution from an innovation and a regulatory science perspective. Multiple occurrence of a particular topic relating to innovation or regulatory science was deemed important and was tagged as a transformation dimension. After review of the articles, the identified transformation dimensions were classified into similar groups termed transformation triggers (2). These triggers are carried forward to this paper for verification from an operational perspective. The goal of this paper is to furnish the corresponding operational evidence that demonstrate likely trends in the pharmaceutical industry with respect to the 14 transformation triggers. The operational evidence is derived from data collected on pharmaceutical companies, products, and technologies. This approach is predicated upon the hypothesis that such data has the potential to provide valuable operational information about the transformation within the industry. The key elements of this approach include description of the methodology used for data collection, graphical presentation of the results, and commentary on the meaning of the results.
2. Methodology
The operational data on pharmaceutical products was extracted from Datamonitor, ClinicalTrials.gov, the Food and Drug Administration (FDA) Orphan Drug database, and a paper published by Wagner et al. (3). The Datamonitor group is a world-leading provider of premium global business information, delivering independent data, analysis, and opinion across many industries including the pharmaceutical and healthcare industry. ClinicalTrials.gov is a registry and results database of federally and privately supported clinical trials conducted in the US and around the world. ClinicalTrials.gov gives information about a trial's purpose, who may participate, locations, and phone numbers for more details. The US FDA Orphan Drug Act provides for granting special status to a product to treat a rare disease or condition. The product, or the combination of products, to treat the rare disease or condition must meet certain criteria. This status is referred to as orphan designation, and drugs designated by the FDA as orphan are searchable in the Orphan Drug database. The paper by Wagner et al. (3) is a global survey of companies pursuing nanomedicine application in the pharmaceutical and medical device industry. At the time of data collection, this paper was the only comprehensive source of nanomedice applications in the pharmaceutical industry.
The data were collected based on the search criteria described in Table I and Table II for 37 top pharmaceutical companies (Table III). Data collection was performed in 2010 but excludes the latter part of 2010. The scope of data collection also included the developmental drug products with a future launch date within the years of 2010 to 2015.
For each of the 14 transformation triggers listed in Table I, the objective and search attributes for collecting operational data was defined. Databases were searched, and data related to pharmaceutical companies, products, and technologies—that is, the operational evidence— were collected. The data were plotted in Figures 1 to 12, and the resulting trends are presented in Section 3. The strength of the operational evidence was determined through visual examination of the figures, and the observed trends were assigned a strength value within the following strength scale to enable a simple comparison with the strength of the theoretical evidence (2). The method for measuring strength of the operational evidence was based on the interval scale with values from 0 to 100. The scale was divided into seven equal intervals with “None” at the lowest end and “Very Strong” at the highest end of the interval (see list below). Seven intervals were chosen to provide maximum precision in allocating the strength values for the operational evidence. Each interval qualitatively defines the amount of operational evidence present for transformation triggers. The “Very Strong” interval was subdivided into two tiers to differentiate between evidence from multiple versus single operational indicator. The word indicator refers to the category of operational data that was used to build the operational evidence (see Section 3). Three indicators were used to build the operational evidence for transformation Trigger 3 (personalized medicine), and hence the operational evidence for this trigger was assigned the highest strength value.
Very Strong (tier 1): Significant amount of operational evidence from multiple indicators exist (Interval Scale 86–100).
Very Strong (tier 2): Significant amount of operational evidence from single indicator exist (Interval Scale 71– 86).
Strong: A reasonable amount of operational evidence exists (Interval Scale 57–71).
Medium: Some operational evidence exists (Interval Scale 43–57).
Weak: Little operational evidence exists (Interval Scale 29–43).
Very Weak: Very little operational evidence exists (Interval Scale 14–29).
None: No operational evidence exists (Interval Scale 0–14).
Trends in operational evidence and the associated theoretical evidence were compared to identify areas of similarity and contrast. The comparison was done by computing the difference between the strength of the theoretical evidence and strength of the operational evidence. The correlation between the two types of evidence was deemed excellent if the computed difference was ≤10, good if 11 to 20, acceptable if 21 to 30, and weak if >30. The computed difference was based on the argument that numerical distance between strength of the theoretical and operational evidence is a simple indicator of their similarity or contrast. The lower values of computed difference mean high similarity in strength between the theoretical and operational evidence, and conversely the higher values of the computed difference indicate low similarity.
3. Results and Discussion
The theoretical evidence for the 14 triggers listed below was established in the article by Shafiei et al. (2). The corresponding operational evidence is established in this paper and the results presented below. The discussion for each trigger reflects the interpretation of the operational evidence as illustrated in Figures 1 to 12. The computed difference between strengths of theoretical and operational evidence, presented in Table IV, suggest an excellent correlation with respect to Trigger 2 (fully integrated pharma network), Trigger 3 (personalized medicine), Trigger 5 (translational medicine), Trigger 7 (regulatory harmonisation), Trigger 8 (science- and risk-based regulation), Trigger 11 (biotechnology), and Trigger 13 (bioinformatics). This means that there is a strong agreement between the theoretical and operational evidence for these triggers. However, the areas of strong contrast where the computed difference is weak are Trigger 1 (healthcare management focus), Trigger 6 (adaptive clinical trials), Trigger 9 (progressive/live licensing), and Trigger 10 (regulatory enforcement). The areas of moderate similarity between the theoretical and operation evidence, where the computed difference is good/acceptable include Trigger 12 (nanomedicine), 14 (pervasive/cloud computing), and Trigger 4 (virtual research and development).
Trigger 1—Healthcare Management Focus
An increase in pharmaceutical revenues from products or services other than from the traditionally strong prescription drug sales would mean that the pharmaceutical industry is diversifying and that Trigger1 is taking root. Revenue information relating to non-prescription drug products of 37 pharmaceutical companies was used as the primary indicator of diversification in the pharmaceutical industry. Since diversification is divergence from established core products/services, of the 37 pharmaceutical companies listed, only those that had “non-prescription drug” and “other” revenue information were selected for trend observation. This limited the final list to 16 pharmaceutical companies. The actual and projected revenue information was collected for financial years 2002 to 2015. Operational trends observed in Figure 1 show a substantial increase in diversified revenue.
Trigger 2—Fully Integrated Pharma Network
The ratio of internally developed versus externally developed drug products is an indication of the degree to which the pharmaceutical industry is leveraging external sources of innovation. To determine this trend the Datamonitor database was searched for products that were launched or to be launched between 2002 and 2015. Sources of launched or to be launched products in the database were categorized as “internal”, “acquired product”, “co-developed”, “M&A”, “other external”, “in-licensed” and for products in development phase “n/a”. The word Internal means that the products were developed in house. Acquired product means that the product was purchased from another organization. Co-developed means the product was co-developed with another pharmaceutical company under specific agreement. M&A means the product was inherited through merger and acquisition. In-licensed product refers to transfer of a license by agreement from another organization in order to develop or market the particular product. Other external refers to acquisition of products externally by other means than explained above. The term “n/a” means not applicable and is used for products in the development phase. The acquired product, co-developed, M&A, in-licensed and other external were collectively consolidated into a single category called “external”. For the purposes of this analysis “n/a” was excluded. The age of the drug product was categorized into very old >15 years, old = 11-15 years, recent = 5-10 years, new<5 years. The prevalence of external sources of products for the newer products would be a positive indication that Trigger 2 is taking root. The trends observed in Figure 2 show a substantial increase in external sourcing for newer products.
Trigger 3—Personalized Medicine
The primary operational indicator for prevalence in personalized medicine is increase in approved Biomarkers which is closely related to discussions on Translational Research (Trigger 5). Other useful but secondary indicators are prevalence in biological products and FDA orphan drug designation. Since personalized medicine, by definition, is concerned with the development of drugs for niche patient populations (2), designation of orphan drugs by the FDA is a good indicator of trends in personalized medicine. Data collection focused on drugs that had received orphan drug designation between 1993 and 2010. The operational evidence for this trigger is supported by three indicators: (i) approved biomarkers, (ii) orphan drug designations, and (iii) biological products. Trends observed shows gradual increase in FDA-approved biomarkers (see Trigger 5), a substantial increase in FDA orphan drug designations (Figure 3), and gradual increase in Launch of biological (large-molecule) drugs (see Trigger 11).
Trigger 4—Virtual Research and Development
A key feature of virtual research and development involves outsourcing research activities to third parties or in some cases co-development with other pharmaceutical companies (2). In order to investigate the likely trends in virtual research and development, the collected data were classified into three categories: (i) drug products developed through internal research and development, or (ii) drug products developed externally through third party agreements, or (iii) drug products developed through partnerships with other pharmaceutical companies. The externalization and collaboration trends for big pharma, mid pharma, Japan pharma, biotech, and generics were derived by calculating the ratio of externally developed drug products to internal drug products and co-developed drug products to internal drug products. The trends observed in Figure 4 show that mid pharma and generics play leading roles in externalization of research and that collaboration among pharmaceutical companies is low in general but slightly more pronounced in big pharma.
Trigger 5—Translational Research
The goal of translational research is to facilitate exchange of information between preclinical scientists and clinical practitioners to implement in vivo measurements that more accurately predict drug effects in humans (2). Prevalence in regulatory approval of biomarkers is a good indication that translational research is increasing. In order to prove this point, a list of approved biomarkers by the FDA (i.e., in vivo measurements) was analyzed to determine the number of products associated with approved biomarkers and date of biomarker approval for trending purposes. As the FDA does not publish explicit approval date for biomarkers, the date of the earliest published research related to the prototypic drugs (drug associated with the label information defining the biomarker context) was used as a surrogate indicator—see the web link in Table I for a list of valid, approved biomarkers published by the FDA. The word valid is described by the FDA as a biomarker that is measured in an analytical test system with well established performance characteristics and for which there is an established scientific framework or body of evidence that elucidates the physiological, toxicological, pharmacological, or clinical significance of the test results. Although sporadic at times, Figure 5 shows a general upward trend in the number of valid biomarkers over the last two decades.
Trigger 6—Adaptive Trials
Information about clinical trials is often maintained in registry and results databases frequently managed by governmental organizations. One such database that is publically available and also contains information on adaptive trials is Clinicaltrials.gov. This database was searched for studies containing the phrase “adaptive design” in Phase I, Phase II, and Phase III clinical trials that were first submitted to the FDA between 2000 and 2010. Figure 6 shows a steady increase in the number of adaptive clinical trials since 2005, and a sharp decline in 2010 is apparent.
Trigger 7—Global Harmonization
Creation and deployment of international guidelines is the direct indication of regulatory and industry commitment to global harmonization. To validate this assertion, the International Conference on Harmonisation (ICH) guidance database was searched for evidence of harmonization relating to safety, efficacy, and quality of drug products. Trends observed in Figure 7 show that the activities on global regulatory harmonization have remained more or less constant during the last 2 decades except for a large spike in 2009 and 2010.
Trigger 8—Science and Risk-based Regulations
Research conducted by regulators in cooperation with the industry and other research organizations was used as a surrogate indicator that regulatory rule making is likely to benefit from results of such cooperation. The FDA database containing a list of cooperative research and development agreements was searched. The review of the FDA's Cooperative Research and Development Agreements (CRADAs) resulted in a classification of research focus into one of the following categories: bioinformatics, personalized medicine, critical path initiative, process analytical technology, biotechnology, pervasive computing, nanomedicine, quality by design, and other categories. The trends observed in Figure 8 show that the agreements are largely focused on bioinformatics, personalized medicine, and in support of FDA's critical path initiative.
Trigger 9—Progressive/Live Licensing
The FDA, European Medicines Agency (EMA), and Health Canada websites were extensively searched for evidence of procedures for drug product licensing that allowed progressive use of medicinal products, that is, starting the commercial use in Phase III clinical development under certain conditions. Although there were some forward-looking statements in the Health Canada website, there was no indication in any of these regulatory websites that medicinal products intended for human use are awarded progressive marketing authorization while in the clinical development phase. There was no operational evidence in support of this transformation trigger.
Trigger 10—Regulatory Enforcement
The issuance of observations by the regulators to pharmaceutical companies is an indication of their enforcement of applicable regulations. Although this takes place in the US, EU, and other regulated markets, due to the Freedom of Information Act in the US only FDA warning letters are available publicly. Trends observed in Figure 9 shows that the issuance of FDA warning letters seemed cyclical since 2000 with a sharp increase in 2009.
Trigger 11—Biotechnology
Trends in commercialization of small molecule drug products (chemical basis) compared with large molecule drug products (biological basis) in the pharmaceutical market can be used as an indicator to determine the position of biotechnology in the pharmaceutical industry. To substantiate this, launch information for small and large molecule drug products for global, US, five EU countries (5EU: France, Germany, Italy, Spain, and the UK), Japan, and the rest of the world markets was extracted form the Datamonitor database and analyzed. Trends observed in Figure 10 show that the number of drug products containing small molecules has risen since 2002 with a sharp decline in 2011. At the same time the number of drug products containing large molecules increased gradually, and the projected convergence with small molecule drug products can be seen by 2014.
Trigger 12—Nanomedicine
Each nanomedicine product listed in the work of Wagner et al. (3) was classified into 9 therapeutic categories (cardiology, central nervous system, genitourinary, immunology and inflammation, infectious diseases, metabolic disorders, musculoskeletal, oncology, and ophthalmology). The number of nanotechnology-based drug products for each therapeutic category was determined. The trends observed in Figure 11 show uneven peaks and troughs in marketing of nanotechnology-based drug products since 1993, with an isolated rise in 2005.
Trigger 13—Bioinformatics
Examining patent information on a particular technology can provide evidence of its prevalence and likely future trends. To test this assertion, the bioinformatics search keywords in Table I were searched in the US and EU patent databases. Trends observed in Figure 11 show a rise in bioinformatics patents since 2000 with peaks at 2004 and 2008. The 2010 data do not represent the full year.
Trigger 14—Pervasive/Cloud Computing
Examining patent information on a particular technology can provide evidence of its prevalence and likely future trends. To test this assertion the pervasive computing search phrases were grouped into five themes of telemedicine, implantable drug delivery, implantable biosensors, intelligent medication package, and remote patient monitoring. These themes were derived from and highlighted by the literature (2) as the possible areas of pharmaceutical applications. The US and EU patent databases were searched according to the search criteria stated in Section 3. The trends observed in Figures 12 and Table V show a substantial rise in number of pervasive computing patents since 2000 with key areas of focus on intelligent medication package and telemedicine. Note that 2010 data does not represent the full year.
Summary Discussion for All Triggers
The operational evidence presented in Section 3 provides substantive evidence in support of pharmaceutical industry transforming from a prescription drug–centric industry to a diversified healthcare industry (Figure 1). Changes in the pharmaceutical business model are also evident in that there is more focus on external sources for supplementing the product portfolio (Figure 2). The newer drug products are three times as likely to be sourced externally as developed internally. The pharmaceutical industry move towards individualized medicine is supported by orphan drug designation (Figure 3), development and availability of valid genomic biomarkers (Figure 5), as well as industry shift from a small molecule blockbuster drug strategy to a large molecule–based targeted drug strategy (Figure 10). For virtual research and development, the operational evidence can be interpreted in two ways: (a) healthy increase in externalization in that the pharmaceutical companies are increasingly exploiting external sources of innovation (Figure 4), and (b) comparatively less enthusiasm on collaborative drug discovery and development among pharmaceutical companies (Figure 4). The modest but steady increase in the number of approved clinical biomarkers (Figure 5) is apparent and is a surrogate indicator that clinicians and scientists are working closely in the context of translational research to develop products tailored for specific populations. The operational data in support of translational research exclude additional evidence which was not known during the initial data collection (August 2010), that is, 14 biomarkers that currently are in the review and consultation stage within the FDA (4). There is enough operational evidence to support the concept of adaptive clinical design (Figure 6); however, it is a small proportion of all the clinical studies that are conducted within the same time period (as of September 17, 2012 there are 132,526 clinical trials with locations in 179 countries according to clinicalstrials.gov). Although regulatory harmonization relating to common safety, efficacy, and quality guidance (1) is firmly supported by the operational evidence (Figure 7), the current data collection found no evidence to suggest that the different regulatory authorities will eventually fully harmonize the pre-market evaluation and post-market surveillance of drug products. Operational evidence for science- and risk-based regulations is mainly limited to efforts of the US FDA's cooperative research agreements and the EMA's innovation taskforce, which are largely focused on personalized medicine, translational research, and bioinformatics topics (Figure 8). Progressive product licensing, although a revolutionary concept, has not been implemented in practice; this was confirmed since at the time of data collection no operational data was found to substantiate this activity. It is likely that this topic will remain in the conceptual phase until there are robust methods to firmly assure product safety at early stages of product development that may be possible in the arena of the personalized medicine. Regulatory enforcement data are only based on the US FDA due to the Freedom of Information Act in the US; enforcement data for EMA were not publicly available during the data collection period. The operational data point to cyclical enforcement profile except a sharp increase in 2009 (Figure 9); this is widely attributed to the FDA commissioner's tough stance on effective regulatory enforcement (5). Application of biotechnology is supported by strong evidence that the projected pharmaceutical product portfolio within the next 5 years will have equal number of large and small molecule drug products (Figure 10). This supports the literature assertion that pharmaceutical industry is focusing more and more on biologics (2), which is also consistent with industry move towards personalized medicine (2). The operational data in support of nanotechnology is somewhat erratic (Figure 11). Clearly there is evidence that nanotechnology plays a role in drug development; however, the amount and consistency of operational data does not indicate a steady rise. Bioinformatics as an enabling technology (2) supporting translational research and personalized medicine is taking root, and its prevalence in the healthcare industry can be noticed in analysis of the worldwide patent data since 2000 (Figure 11). The operational evidence supports the literature assertion that pervasive computing will increasingly play a key enabling role in the pharmaceutical industry with a particular focus on patient support activities such as intelligent medication, telemedicine, and remote patient monitoring (Table V, Figure 12).
4. Conclusion
In general there is a good correlation between the theoretical evidence derived from the literature review and the corresponding operational evidence. The strength of the operational evidence supports the literature findings that Trigger 2 (fully integrated pharma network), Trigger 3 (personalized medicine), Trigger 5 (translational research), and Trigger 14 (pervasive computing) are important drivers of pharmaceutical industry transformation (Table IV) and hence should be used to determine their impact on pharmaceutical quality. Key areas of contrast, however, are Trigger 1 (healthcare management focus), Trigger 6 (adaptive/in-life trials), and Trigger 10 (enforcement) for which the operational evidence is stronger than the theoretical evidence (Table IV). A general explanation for this contrast could be attributed to the fact that there is a paucity of academic research in the field of pharmaceutical quality (2). This is particularly true for Trigger 10 (enforcement) and to a lesser extent to Trigger 1 (healthcare management focus). The explanation for Trigger 6 (adaptive/in-life trials) is more nuanced and could be linked to the originator of the adaptive trial concept (i.e., led by industry, hence, lag in academic work).
The theoretical evidence in Reference 2 and the supporting operational evidence presented in this paper were examined through a questionnaire-based opinion survey in order to verify the findings from the perspective of the experts in the field (6).
Limitations
To the extent possible, both FDA and EMA databases were consulted for regulatory evidence. However, in some cases (e.g., regulatory enforcement), due to the Freedom of Information Act the only publicly available data was found in the US FDA databases. Regardless of country of origin since the US FDA has an extensive regulatory network and global oversight on drug products marketed in the US, this limitation does not negatively affect the validity of the results.
Conflict of Interest Declaration
The authors declare that they do not have any financial or nonfinancial competing interests related to the content of the manuscript.
Acknowledgements
The corresponding author works for Sanofi (Bridgewater, NJ) and gratefully acknowledges the company's support for this research as part of his personal development. Views expressed in this paper are those of the authors and do not in any way reflect the official policy or position of Sanofi.
- © PDA, Inc. 2013
PDA members receive access to all articles published in the current year and previous volume year. Institutional subscribers received access to all content. Log in below to receive access to this article if you are either of these.
If you are neither or you are a PDA member trying to access an article outside of your membership license, then you must purchase access to this article (below). If you do not have a username or password for JPST, you will be required to create an account prior to purchasing.
Full issue PDFs are for PDA members only.
Note to pda.org users
The PDA and PDA bookstore websites (www.pda.org and www.pda.org/bookstore) are separate websites from the PDA JPST website. When you first join PDA, your initial UserID and Password are sent to HighWirePress to create your PDA JPST account. Subsequent UserrID and Password changes required at the PDA websites will not pass on to PDA JPST and vice versa. If you forget your PDA JPST UserID and/or Password, you can request help to retrieve UserID and reset Password below.