Why does in-licensing remain a popular strategic option for many pharmaceutical companies if, as is repeatedly claimed, in-licensed drug candidates are more likely to fail than succeed? Much of the available research emphasizes their comparatively high failure rates. Pavlou and Belsey1, for example, recently surveyed pharmaceutical industry executives who observed failure rates of 21–50% for licensing deals. PriceWaterhouseCoopers in Europe estimated2 that 59% of product development alliances for large pharmaceutical companies failed to meet expectations. Similarly, more broadly defined studies by consultancy companies or academics have usually concluded that 40–60% of alliances disappoint3. These and a plethora of similar studies give pharmaceutical companies good reason to be concerned about their ability to successfully manage joint discovery projects. The question we ask here is whether these deal breakdown rates and the inability of many alliances to live up to expectations should be reason for specific concern when drug development is generally harassed by unavoidable attrition.
Success rates for in-licensing
Facett armchair, courtesy of Ligne Roset, www.ligne-roset.co.uk, 0870 7777202Contrary to widespread belief, there is evidence that the development of in-licensed compounds is not necessarily more prone to error than that of self-originated compounds. The Adis International R&D Insight Database, for example, contains close to 2,000 compounds in development in more than 900 biotechnology and pharmaceutical companies in the United States. Approximately 50% of these compounds were developed in an alliance at the clinical development stage. In fact, studies based on this and comparable data sources (PHID, SCRIP, Tufts and Centre for Medicines Research International Institute for Regulatory Science) have consistently shown a higher probability of reaching the market for in-licensed compounds than for self-originated compounds (fig. 1).
Figure 1 | Probability of success from Investigational New Drug approval to marketing approval in different studies.
DiMasi's data5 covers 671 new chemical entities developed by 24 large pharmaceutical companies in the period 1982–1991. Gambardella et al.6 use the PHID database which covers 2,078 projects entering Phase I in the period 1989–1999. The Centre for Medicines Research (CMR) data7 is based on 432 self-originated and 86 licensed-in new active substances developed in 33 companies that entered a clinical phase during 1998– 2000, where a decision was made by 31 December 2003. The percentages reflect first human dose to market success probabilities.
A closer inspection of CMR data reveals that the probability of success is higher for in-licensed compounds in the early clinical stages, and lower in Phase III and in Submission (fig. 2). This will tend to increase the resources spent in vain on in-licensed compounds above the level for self-originated compounds, which will tend to fail earlier. Still, the markedly higher overall success rates for in-licensed compounds makes in-licensing an attractive option.
Figure 2 | Probability of success in each of the clinical phases and at submission.
Based on the Centre for Medicines Research (CMR) data7 on New Active Substances as outlined in fig. 1 and found in the CMR International 2005/2006 Pharmaceutical R&D Factbook. Used with permission from CMR.
Challenges with in-licensing
We interviewed senior executives from three of the top ten European pharmaceutical companies to discover their reasons for the higher probability of success for in-licensed compounds. The executives were either managing the licensing processes, the development of in-licensed and self-originated compounds, or they were represented at the committees governing the development projects. They were interviewed about their experience from projects that they had been personally involved in, and about their general impression of the subject matter for the industry.
The interviewees all agreed that the development of licensed compounds poses additional challenges to a pharmaceutical company. This is partly due to the need for evaluating in-licensing opportunities over only a few weeks and based on information provided by an external organization. The challenge is generally considered manageable, though, as much of the information about the compound provided by the potential partner adheres to regulatory demands and guidelines, is reasonably objective and can be easily compared to results for in-house projects. This is especially true for information regarding the compound's safety and efficacy, which is crucial in determining the potential of the drug in therapeutic use. A second challenge stems from the significant transfer of knowledge about the compound that will have to take place from the biotechnology company to the pharmaceutical. Finally, it might also be the case that the biotechnology company will be actively involved in the product's development, which again poses additional challenges in terms of coordination and the management of conflicts of interest.
Key factors for success
figure 3 outlines the executives' suggested reasons why there would still be a higher rate of success for in-licensed compounds than for self-originated compounds. The factors can be summarized as follows: in-licensed projects need a relatively higher expected pay-off; and competition created by in-licensed compounds means that less promising self-originated compounds can be preferentially pushed through development or put on the backburner. Overall, those interviewed did not consider their in-licensed projects to be any easier than self-originated projects, suggesting that higher success rates cannot easily be attributed to a subset of 'easy' projects having been selected for in-licensing. On the contrary, in-licensed compounds would often be developed from novel concepts and more likely to be outside the pharmaceutical's core research competencies than in-house originated projects.
Figure 3 | Why in-licensed compounds can have a higher rate of success than self-originated compounds.
Comments are summaries of interviews with senior executives from three of the top 10 European pharmaceutical companies.
Recommendations
In-licensing can pose many challenges within R&D divisions in big pharmaceuticals, but those interviewed all emphasized the need to continue building relationship-management capabilities. Some argued that their company should pay more attention to difficulties raised by biotechnology partners being responsible for major work streams in the development process. A main reason why this situation poses additional challenges is that the biotechnology companies are often less risk-averse and less attentive to the need for standardization and documentation of processes, which will tend to reduce the probability that regulatory approval of the product is granted.
It is very likely that the future will continue to see the growth of in-licensing in pharmaceutical companies. As recently argued by Jones and Clifford4, in-licensing is an attractive option for pharmaceutical companies because it increases flexibility and provides additional growth opportunities. At the same time, the amount spent in the biotechnology industry — and therefore the number of potential compounds coming from the industry — is growing, and this tends to increase the chance that pharmaceutical companies will find very attractive in-licensing opportunities. The added challenges in developing licensed compounds could still limit demand. Yet when it becomes generally recognized that in-licensed compounds fare better than expected throughout product development, this demand could well rise.
