Contract cell line development (CLD) is a service provided to Bio/Pharma, in which cell lines are engineered into bioproduction systems that stably express recombinant, therapeutic proteins. At the time that I was looking into this topic I was first starting to realize the disciplinary siloes to which analyses (at least publically available ones) generally restrict themselves (especially technical vs business discussions). I decided to dig deeper into CLD to consider and connect commercial, technical, and psychological perspectives. The result was probably a little too academic. Nevertheless I share my train of thought below.
Table of contents:
Economic rationale for Bio/Pharma outsourcing
Technical requirements of cell line development
The major market for cell line development
Client decision factors for contract cell line development
(1) Reputation
(2) Time
(3) Cost
(4) Technology performance
(5) Service integration
(6) Customer Service
Conclusion: The psychology of risk perception
Economic rationale for Bio/Pharma outsourcing
Economists have traditionally considered the service sector as annexed to increasingly affluent societies, a view that focuses on consumer services [4]. However, consumer services have remained largely stable and growth in the service sector has been driven by business services [2]. The scope of this sector has broadened such that virtually any “value-adding” activity can be hired as an external service.
This “rise of the service economy” can be rationalised through three mainstream theories of the firm: firstly Property Rights Theory, which describes the costs of asset ownership and the consequent need for efficient asset use; secondly the Resource-Based View [1], which describes the limited nature of firm resources and the consequent need for focusing them on core activities; and thirdly Entrepreneurial Theory [2], which describes the importance of dynamically redesigning organisational boundaries around promising business opportunities [3].
These theories suggest that service providers generate value by providing assets, expertise, and/or processes that have lost a central position in the agenda of the client firm. This both frees internal resources of the client firm for strategic focus on its key activities and provides the client firm flexibility for the dynamic pursuit of opportunities by allowing spontaneous access to external resources.
Therefore as previously vertically-integrated industries mature and their processes become standardised to “best practice”, competition forces companies to focus on core competencies and outsource non-core activities, enhancing productivity by division of labour [3]. This is the exact phenomenon that has transformed the biopharmaceutical industry, which is characterised by unusually long, costly, and expertise-intensive product development cycles [3]. Over the past decade and a half, the biopharmaceutical value chain has progressively disintegrated into a highly distributed ecosystem of highly-specialized product developers and service providers [4].
These service providers offer lower costs, faster time-to-market, access to external technologies and specialised expertise, and access to capacity - all for less capital investment and fewer fixed costs [5],[6],[7]. Consequently, outsourcing has become a popular strategy with large Bio/Pharma for cost-containment and supply chain flexibility, whereas SME Biotech utilise these vendors to access vital product development processes. As shown in Figure 1, these include discovery services, lead optimisation, clinical trial management, contract manufacturing, and more.
This article focuses on CLD, which is performed during the preclinical phase with the aim of establishing a stable and scalable system to generate GMP product material at animal-trial, clinical-trial, and commercial scales [8]. CLD therefore immediately precedes the key value accrual stages of IND filing and clinical studies.
Technical requirements of cell line development
In CLD, recombinant genes expressing the biopharmaceutical product are stably transfected into a host cell line for bioproduction. The transfected cell population is then serially diluted to obtain the highest performing clonal line, where such monoclonality is a regulatory requirement to help ensure good batch-to-batch consistency. Batch-to-batch consistency minimises the risk of variable immunogenicity and should therefore be a prime performance measure.
Passaging stability is another important performance measure [9], affecting scalability. For example cell lines must divide for 60-70 generations to scale from the working cell bank to a 2-3000 litre bioreactor. Client firms must therefore consider passaging stability in relation to the size of the target market, as well as a third vital performance measure: volumetric productivity [10]. This determines the production scale required for a given output of product material and therefore significantly impacts cost-effective scaling and biomanufacturing COGS. Thus these three performance characteristics, which are also significantly optimised in later process development [11], are selected for during CLD.
The client’s choice of cell line will be product-dependent and may require different vendor infrastructure, technology, and expertise. For example monoclonal antibodies (mAbs) are characterised by complex post-translational modifications that require mammalian cell line hosts, of which the dominant host is Chinese Hamster Ovary (CHO) [12],[13]. CHO has a long track record from the licensure of tissue plasminogen activator in 1987 [14] and currently accounts for ~70% of industrial mammalian biomanufacture [17]. From a technology perspective therefore, mammalian CLD is highly consolidated.
The major market for cell line development
Mammalian CLD presents an attractive market, not only because it commands significantly greater prices than microbial CLD [15], but also due to the dominance of the biopharmaceutical market by mAbs, which accounted for 37% (338 of 907) of all biopharmaceuticals in development in 2013 [16]. As 54% of biopharmaceutical companies outsource CLD [17], it can be estimated that 20% of all biopharmaceutical developers in the world outsource mammalian CLD for mAb development. Moreover at 12% CAGR, global mAb sales are forecasted to grow faster than both small molecules and other biopharmaceuticals [18] to reach $171 Bn by 2017 [19]. Contract CLD vendors therefore focus on mammalian CLD.
The strong growth of the mAb market and of biopharmaceuticals generally provide incentive to better understand associated contract services, particularly CLD, which seems largely ignored by multidisciplinary research.
Client decision factors for contract cell line development
What factors influence the selection of contract CLD vendors by Bio/Pharma clients? To find out I interviewed 30+ professionals from CLD vendors, large Bio/Pharma clients, and SME Biotech clients.
Delving into this I found 6 main decision factors: (1) Company reputation, (2) Time, (3) Cost, (4) Technology performance, (5) Service integration, and (6) Customer service. It seems that risk aversion dominates the evaluation and prioritization of these decision factors. Risk aversion captures the market around “branded” CLD vendors; and acts as a significant barrier to innovation in bioproduction cell lines. I would speculate that risk perception can be modified by specific trust and control strategies. There also seems to be a under-recognized request for better customer service by SME Biotech, which is potentially an opportunity for unique engagement strategies.
(1) Reputation:
Company reputation includes the company’s track record of providing cell lines whose product successfully enters clinical trials. Reputation is considered the top selection criterion [22]. Hence clients have notable power to damage reputation by communicating dissatisfaction, and vendors have the liability thereof. Despite not explicitly stated as such by respondents, reputation is a surrogate measure for low risk; such risk aversion also manifesting in conservativeness to novel technology.
(2) Time:
Minimising project time is important to accelerate time-to-market and thus avoid wasting patent coverage of the product. Moreover it is critical to smaller companies due to their cash burn. The average turnaround for a CLD project quoted by SME Biotech and large Bio/Pharma, 7.5 months and 8.2 months respectively, were not significantly different. Many CLD vendors add a 50% contingency to account for technical difficulties.
These technical difficulties are endemic, with potential repercussions to the reputation aspect. When presented with a scenario choice between a quick project turnaround without GLP or a slower turnaround under GLP (rationalised by the associated documentation burden), 92% (n=24) chose the latter despite being reminded GLP is not a regulatory requirement. This observation is surprising for SME Biotech given their cash burn, but can be rationalised through risk-aversion. Indeed CLD vendors indicate that spending time on a quality cell line can massively mitigate later development difficulties, as well as manufacturing COGS.
(3) Cost:
Project cost is represented by both upfront price and licensing structure, the balance of which is determined by resource availability for upfront payment, as well as the interest in accessing proprietary technology. At £154k - £221k (SE ± £19k) and £213k - £288k (SE ± £31k) respectively, the average range quoted by SME Biotech was lower than for large Bio/Pharma although without statistical significance. The ability to compete by undercutting these standards was investigated by asking respondents for what discount on their current CLD costs they would switch vendors.
All responding large Bio/Pharma (n=4) and 56% of SME Biotech (n=9) indicated that price discounting was not a sufficient incentive to switch, whereas the remaining SME Biotech (n=7) reported an average discount of 19% (SE ± 2%). Therefore although undercutting may work with the most price conscious SME Biotech, the upfront price component of the cost aspect is not very effective in influencing vendor selection. In investigating the converse, respondents were asked what factors justify a higher price. They answered reputation and technology performance, with minor mention of one-stop-shop, customer service and favourable licensing terms.
For the licensing component of cost, 50% (n=16) of respondents indicate having accepted licensing terms for CLD. They suggest that licensing structures are common, and highlight that established CLD vendors often use conditional licensing structures to incentivise the use of their biomanufacturing services, leading to limited flexibility in downstream CMO choice. However respondents expect that licensing structures should fade out for standard technology, especially with less established CLD vendors.
(4) Technology performance:
Technology performance, which or may not be proprietary, is most crucial in relation to enabling the biomanufacture of product in a manner both cost-effectively scalable and satisfactory to stringent regulations. I saw emphasis largely on the latter objective, failure of which necessitates repetition of CLD. For this reason, significant preference exists for well-validated technology over novel, higher-performance technology.
Prioritisation of batch-to-batch variation over volumetric productivity corroborates the emphasis on regulatory acceptability over cost-effective scalability. Furthermore, large Bio/Pharma exhibit a significantly greater concern for passaging stability. This may be explained by their frequent involvement in commercial scale biomanufacture, in contrast to SME Biotech whose exit horizon typically limits them to clinical trial-scale biomanufacture at which less passaging occurs. Therefore within the technology aspect, performance measures can be prioritised based on the firm’s product development horizon.
(5) Service integration:
Service integration is the concept of the one-stop-shop vendor. The interaction between CLD and adjacent activities can enable time-saving and simplification through service integration, for example process and analytic development may both draw from and feedback to CLD. The most significant advantage may be to avoid the problematic process of technology transfer as well as of managing multiple vendors.
Within reason, service integration is prioritized over upfront cost to avoid unnecessary technology transfer and multi-vendor management; both of which also contribute to time saving. Pilot- and clinical trial-scale biomanufacture command significantly more integrated services premium than commercial-scale biomanufacture, particularly for SME Biotech for whom the latter falls beyond the product development horizon. Thus as for technology performance, the product development horizon may impact prioritisation of service integration. For SME Biotech, commercial-scale biomanufacture and fill/finish are too far downstream from CLD to effectively influence vendor selection.
Notably, although outsourcing biomanufacturing continues to be viewed favourably by large Bio/Pharma, the outsourcing of CLD is perceived less so. It has been realised that the valuable product and manufacturing insights derived from CLD are not sufficiently generic to be easily captured from vendors. Moreover, competitive intelligence concerns arising from the biosimilars industry (for whom a significant barrier is the non-generic nature of cellular manufacture systems) may also deter outsourcing CLD - particularly as several CDMOs are diversifying into biosimilars.
(6) Customer Service:
Respondents were asked to indicate the greatest difficulties and risks of contracting CLD. Customer service appeared in multiple guises including as a lack of transparency and effective communication - the dominant vendor difficulty. For large Bio/Pharma lack of transparency related to difficulty uncovering and troubleshooting problems, but for SME Biotech the following narrative emerges:
SME Biotech firms have few staff as well as few product candidates, which are therefore vital to the long-term viability of the firm. Consequently the executive team tend to micromanage development, especially as they report directly to the Board. When outsourcing CLD, developmental control must be surrendered but executive staff desire frequent updates, including in-depth discussions of data rather than summaries.
In conflict with this, many CLD vendors operate one-stop-shop models including biomanufacture capabilities. These integrated vendors reportedly prioritise large Bio/Pharma clients, who are more likely to provide commercial-scale project-work than SME Biotech. Moreover, vendors catering to Bio/Pharma are unused to “collaborative” product development, in part because Bio/Pharma have the internal expertise to specify their exact requirements and to help troubleshoot technical issues, and in part because large vendors operate set processes with little scope for bespoke services/development.
By contrast, SME Biotech must rely much more on vendor expertise. In parallel, successful product development will require drawing on both vendor and client expertise to troubleshoot technical issues. Therefore the SME Biotech/vendor paradigm requires greater collaboration, SME Biotech executive staff desire greater transparency and interaction, yet may receive “B-team” or otherwise unsatisfactory service from integrated vendors focused on large Bio/Pharma.
In order to investigate the relative importance of this novel aspect, I offered respondents a scenario choice between a one-stop-shop vendor (from CLD to GMP biomanufacture) with little transparency/collaboration but a well-established brand, or a CLD-only vendor with transparent/collaborative project management. 33% of large Bio/Pharma (n=4) vs 79% of SME Biotech (n=14) chose the CLD-only vendor. Thus customer service is an aspect that can effectively compete with service integration, such that Reputation>Technology=Customer Service>One-stop-shop=Time>Cost.
Customer service was not mentioned as a key aspect by client or CLD vendor respondents. It appears to be under-recognised and therefore potentially an opportunity for unique engagement strategies.
Conclusion: The psychology of risk perception
Risk aversion dominates selection-decisions of CLD vendors. Risk aversion leads clients to select by reputation rather than trial (reputation factor), to select well-validated technology over higher performance (technology factor), to emphasise technology characteristics concerned with regulatory risk rather than cost or scalability (technology and cost factors), to sacrifice project time for the application of GLP absent regulatory requirement (time factor), and to accept licensing structures for access to validated technology or companies (cost factor).
Whereas contract CLD vendors are highly open to novel technology, their clients are not. SME Biotech cannot afford the risk of trailblazing, and although large Bio/Pharma have the resources to trial innovative technology in parallel to traditional technology, they are not motivated. Thus the mantra “the process defines the product” espoused by the biologics industry is here applied to avoid changing the process rather than to leverage it for competitive advantage, a result of perceived regulatory risk on technology performance considerations.
Alliance strategy literature suggests that perceived risk may be divided into relational risk, i.e. the apprehension of unsatisfactory cooperation such as opportunistic behaviour; and performance risk, i.e. the apprehension of failing to achieve objectives. Here it seems performance risk is the principal dimension of client firm concern, due to the technically difficult and unpredictable nature of biopharmaceutical development. However literature also suggests that younger and smaller contract manufacturers intrinsically operate at a higher level of quality risk, in part due to relational risk and in part because incentives are misaligned with those of the client firm. Therefore CLD vendors, including those integrating biomanufacturing services, must address both relational and performance risk when engaging with client firms.
To this end, control and trust are principal aspects that may be leveraged to influence perceived risk. Trust may be divided into goodwill trust and competence trust, which mitigate the perception of relational risk and performance risk respectively. Regarding control, only output control is pertinent to this discussion and relates to implementing measures of performance satisfaction that mitigate performance risk.
Reputation, which is here reported the dominant selection aspect, mitigates both relational risk and performance risk by inspiring goodwill trust and competence trust respectively. A CLD vendor lacking reputation must devise alternative strategies to mitigate perceived risk. Some recommendations could include providing transparent customer service, which mitigates relational risk by reducing knowledge asymmetry; as well as providing collaborative customer service, which mitigates relational risk by building goodwill trust, and also mitigates performance risk by offering the client a vehicle for indirectly controlling the CLD vendor’s activities.
Notably, the distinction between actual and illusory control is not significant to the mitigation of perceived risk. Therefore the practicality of extending client control to the activities of the CLD vendor is not necessarily an issue, for example the impression thereof may be sufficient to satisfy the greater micromanagement tendencies of SME Biotech.
[i] Das et al. (1996), “Risk types and inter-firm alliance structures”, Journal of Management Studies 33: 827
[ii] Gray et al. (2012), “Quality Risk and Contract Manufacturing: Theory and Empirical Evidence”, Working paper, Tuck at Dartmouth
[iii] Das et al. (2001), “Trust, Control, and Risk in Strategic Alliances: An Integrated Framework”, Organization Studies 22: 251
[1] Kraaijenbrink et al. (2010), “The Resource-Based View: A Review and Assessment of Its Critiques”, Journal of Management Vol. 36:1
[2] Klein et al. (2012) “Strategic Entrepreneurship”, Strategic management Working Paper 3, Department of Strategic Management and Globalization Copenhagen Business School
[3] Kaitin (2010), “Deconstructing the Drug Development Process: The New Face of Innovation”, Clin Pharmacol Ther. 87:3
[4] Sabatier et al. (2010), “From Recipe to Dinner: Business Model Portfolios in the European Biopharmaceutical Industry”, Vol. 43:2
[5] Lowe et al. (2010) “Contract Manufacturing in Life Sciences The Need for an Integrated Approach” Lodestone [online] available at: http://www.lodestonemc.com/files/pdf/WP_Contract_Manufacturing_in_Life_Sciences.pdf (last accessed 30/05/2014)
[6] Gray et al. (2012), “Quality Risk and Contract Manufacturing: Theory and Empirical Evidence”, Working paper, Tuck at Dartmouth [online], available at http://faculty.tuck.dartmouth.edu/images/uploads/faculty/brian-tomlin/brian_tomlin_cv.pdf (last accessed 28/05/2014)
[7] Paragon Sciences (2011), “Global Contract Manufacturing Companies : Pharmaceutical and biotechnology” Special Report [online], available at ttp://www.paragonbioservices.com/LinkClick.aspx?fileticket=a4tkUko4CZo= (last accessed 05/06/2014)
[8] Li et al. (2010), “Cell culture processes for monoclonal antibody production”, mAbs 2:5
[9] Estes et al. (2014), “Mammalian Cell Line Developments in Speed and Efficiency”, Adv Biochem Eng Biotechnol 139:11
[10] Lai et al. (2013), “Advances in Mammalian Cell Line Development Technologies for Recombinant Protein Production”, Pharmaceuticals Vol. 6:579
[11] Flueck-Kabay et al. (2014), “Data-Driven Cell-Line and Process Development“, Genetic Engineering & Biotechnology News Vol. 34:4
[12] Kim JY et al. (2012), “CHO cells in biotechnology for production of recombinant proteins: current state and further potential” Appl Microbiol Biotechnol. Vol. 93:3
[13] Fan et al. (2012), “Improving the efficiency of CHO cell line generation using glutamine synthetase gene knockout cells”, Biotechnol Bioeng. Vol. 109:4
[14] Kelley (2009), “Industrialization of mAb production technology: The bioprocessing industry at a crossroads”, MAbs. Vol. 1:5
[15]Cheng Hou et al. (2010), “New frontiers in cell line development: challenges for biosimilars”, J Chem Technol Biotechnol 2011; 86: 895
[16] Pharmaceutical Research and Manufacturers of America (2013), “Medicines in development: Biologics” [online], available at: http://www.phrma.org/sites/default/files/pdf/biologics2013.pdf (last accessed 09/06/2014)
[17] Fowler et al. (2013), “Competing in the Global Contract Manufacturing Market”, Frost and Sullivan report.
[18] Aiswariya (2014), “Technology trends and strategic insights in biopharmaceutical contract manufacturing”, Frost and Sullivan report.
[19] Research and Markets (2014), “Research and Markets: Global and Chinese Monoclonal Antibody Industry Report” [online], available at: http://www.researchandmarkets.com/research/95vpvf/global_and (last accessed 09/06/2014)
Comments