Rational reform of new drug research and development

[China Pharmaceutical network industry trends] China's new drug R & D is in its infancy Because most local pharmaceutical enterprises are still in the stage of "imitation" and "follow-up", and the property in the R & D pipeline is not rich, it must not be a good way to copy the strategy of transnational pharmaceutical enterprises For quite a long time, pipeline R & D, milestone objectives and parallel tasks may still be the best way for local enterprises to improve their rates (rational reform of new drug research and development: transforming the medical alternative pipeline mode is the general trend Picture source: Baidu picture) 1 In recent decades, the inverse Moore's law and productivity crisis, almost all the important technologies involved in new drug research and development, their efficiency has been improved exponentially DNA sequencing for identifying new targets is a billion times faster than that in the 1970s; combinatorial chemistry has increased the number of molecules synthesized by chemists every year by 800 times than that in the 1980s; and the calculation of three-dimensional structure of proteins is three orders of magnitude less than that 50 years ago In view of this, the productivity of drug research and development should have been at a historical high But the embarrassing reality for the industry is that at the same time, drug research and development is facing a serious "productivity crisis": candidate drugs are more likely to fail in clinical trials than in the 1970s; From 1950 to 2010, the R & D cost of a new drug approved doubled every 9 years, a trend that has been very stable for 60 years.. The dilemma of new drug development is known as "eRoom's law" (reverse spelling of Moore's law, which means that the performance of computer chips doubles every 18 months and the price drops by half) Since the 1990s, some scholars have put forward the possible explanation of the inverse Moore's law, which can be roughly divided into three guesses: (1) the hypothesis of drooping fruit: the drugs previously discovered are easy to be found, and the rest are difficult; (2) the hypothesis of regulatory obstacles: after the "reaction stop" event, the regulatory authorities' requirements for new drug application are increasing, causing certain obstacles; (3) research and development mode The first two facts, whether objective or not, are hard to change So, what is the problem of R & D model? 2 With the development of high-throughput screening and combinatorial chemistry in the 1980s and 1990s, and the view that various "omics" can bring a lot of new targets, pharmaceutical enterprises have set off a wave of change to the "industrial assembly line" R & D model During this period, some management strategies practiced by the mature automobile industry, such as Toyota's "Lean Six Sigma" method, seemed to inspire the pharmaceutical industry Many companies set out to improve the efficiency of drug discovery through similar strategies The core points of these R & D management strategies include: strict stage division, overall process optimization, shortening cycle, controlling variation, eliminating delay, etc However, is this development strategy really suitable for new drug research and development? Let's see the following paradoxes: 2.1 milestone goal paradox because the research and development process is strictly divided into different stages, setting the number of goals to reach a milestone in the stage seems to be a direct way of efficiency management The theoretical basis behind this model is rather crude: if one of every 10 projects entering the clinical trial will succeed, then the number of candidates entering the clinical trial will double, and finally two drugs will be listed It is worth noting that there may be an embarrassing fact behind this: in order to achieve the milestone goal, R & D personnel and management inevitably need to consider meeting the quota, and take some bad projects to the next R & D stage This "quantity" driving mode actually damages the quality and sustainability of R & D pipeline 2.2 parallel task paradox another important trend in the 1990s is to divide each stage of R & D into several parallel tasks and strive to shorten each cycle First, it can be expected to shorten the whole cycle, and second, to extend the protection period to a large extent In fact, the development of parallel tasks has seriously damaged the opportunity of rapid decision-making in each stage This not only increases meaningless R & D activities, but also makes it take longer to terminate an unsuccessful molecule There are clear data showing that for the R & D industry with high risk and high loss rate, the sequential strategy has more advantages than the parallel task 2.3 back up paradox in order to avoid possible failure in the later stage, researchers often prepare several backup ups with similar structure for a candidate drug, so as to avoid the longer time spent from the beginning In fact, the current data shows that the strategy of backup basically does not help the success rate, but consumes resources Because of the similar structure, mechanism and PK between backup-up and candidate drugs, it is basically a dead end 2.4 lessons learned 2.4.1 "Innovation Studio" or "assembly line factory"? Assembly line R & D mode has been repeatedly questioned in the practice of drug R & D, and more and more institutions realize that drug R & D has its own special laws, so it is necessary to find a mode suitable for drug R & D itself For example, to develop a new car, a 0.5-kilometer reduction in fuel consumption can be defined as success, while the result of new drug development is often "0" or "100" If R & D productivity is divided into two aspects: efficiency (time and cost of each stage) and efficiency (quality and possibility of success), it is clear that the "industrialized" R & D model only solves the former (efficiency), but may drag the latter (efficiency) The pipeline model is more suitable for predictable and controllable R & D activities For drug R & D, reflection, accidental discovery and creativity are all important factors These two R & D models are also divided into "assembly line factory" and "Innovation Studio", whose driving forces are process control and creativity respectively 2.4.2 in the past ten years, rational reform and mode transformation, the mode of drug research and development began to have some exploratory attempts, and gradually turned into an important strategy A key factor in the decline in R & D productivity is the high rate of loss in phase II and phase III clinical studies The key to improving R & D efficiency is to allocate R & D resources to drugs that are likely to succeed Until a bad project fails in the next stage, it will occupy a lot of R & D resources and block the R & D pipeline Based on this, the "quick kill" strategy is regarded as a panacea to improve R & D productivity The earlier you fail, the less time and resources you save At the same time, the concept of translational medicine was really included in the lexicon of new drug development Using the method of translational medicine, linking the biological basis, preclinical data, biomarkers and clinical research of diseases, and scientifically identifying good projects are recognized as important strategies 3 Pharmaceutical giant's weapon 3.1 Lily: Chorus model 4 principles (4 principles) Lilly may be an actor of this enlightenment In 2002, Lilly secretly set up a department called chorus The team was very small and low-key, hoping that chorus could freely develop a completely different transformation and early clinical research strategy After the establishment of chorus, one of the first tasks is to quickly and cheaply determine the priority of candidate drugs in the R & D pipeline An important tool of chorus is the "reduced proof of concept test (l2poc)", that is, to identify the potential risks as early as possible with low cost The basic logic of chorus is: if a project has a clear target and believes that the probability of success is high, traditional R & D mode can be used step by step; if the probability of success is low, l2poc experiment should be designed as early as possible to judge the risk and reduce the investment Although after more than ten years of development, the principle of chorus has remained unchanged: (1) using the "killer experiment" to deal with the factors that have a great impact on uncertainty as early as possible; (2) treating R & D projects fairly, avoiding the projects forced on the horse due to interest disputes, market value management and capital market expectations; (3) until the key risks of the candidates are removed, the R & D activities in the downstream stage shall be postponed to limit the so-called parallel progress; (4) the organization shall be flat to avoid the delay and bureaucratic wrangles caused by the hierarchical reporting Researchers are allowed to sign contracts independently on outsourcing Clear data have shown that since its establishment, chorus has significantly improved its R & D efficiency, and its founding principles have also been tested by time In 2008, Lilly announced chorus to the outside world and defined it as the future R & D path of Lilly The impact of chorus on the industry is also immeasurable Some people even say that the current booming business model of celgene can be seen as an enhanced version of chorus 3.2 AstraZeneca: five right (5R) AstraZeneca set up an investigation team in 2011 to conduct a detailed retrospective analysis of its R & D pipeline and review the root cause of the candidate logistics loss strategically In their review, five key technical determinants (5R) for the success of the project are summarized, namely: (1) correct target: full understanding of the biological basis of the disease; (2) correct tissue: drug exposure and pharmacological activity in the target organ; (3) correct safety: reasonable safety window; (4) the right patient: select the right patient population for clinical research; (5) the right business potential: the consistency of medical value and business insight AstraZeneca began to vigorously modify its R & D operation strategy in 2011, and the implementation of the 5R strategy directly led to AstraZeneca greatly reducing the number of candidate drugs in early development Each newly started project also clearly analyzes each component of the 5R framework AstraZeneca believes that 5R can help the team to make good decisions at the right stage, while also clarifying the advantages of each project and the transparency of decision-making 3.3 Pfizer: "the three pillars of survival" (the three pillars) were reviewed almost at the same time as AstraZeneca Pfizer also reviewed 44 projects terminated in clinical II The conclusion is that the lack of efficacy is an important factor in the loss of candidates in clinical phase II, accounting for 43% Based on this, Pfizer proposed the concept of "three pillars of survival", namely: (1) PK / PD behavior fully exposed in the action site; (2) proving that the body can combine with the target; (3) showing functional pharmacological activity Pfizer believes that it is high probability that the project with these three elements will be transformed in later clinical trials 3.4 Roche: implement the seven steps of "quick killing" In order to eliminate the low-quality projects as early as possible, Roche proposed a seven step plan for implementing the "quick killing" strategy, including: (1) abandon the denial Blindly pursuing growth and shortening the cycle may bring huge opportunity cost, and the R & D team will smoothly promote those projects that will eventually fail (2) discard "success" and "failure" from the company's dictionary, no matter whether the project is promoted or terminated, a good decision is a success; (3) rebuild the reward system, reward a good decision rather than a good result; (4) abandon the progress goal of seeking "quantity"; (5) quantify the opportunity cost, and treat the decision to terminate the project as expected revenue rather than loss