The pharmaceutical industry is well known for being one of the most inefficient industries globally. Numerous studies, including Steve Paul’s famous 2010 paper, the Tufts Cost study, and multiple landmark papers from the St. Gallen Consortium led by Oliver Gassmann and Alexander Schuhmacher, have provided detailed productivity analyses of internal pharma R&D and collaborations.
Estimates range from $2.8 billion per approval by Tufts (2016) industry-wide to staggering $6.1 billion per approval within Big Pharma according to the recent St. Gallen consortium paper. It still takes over a decade to discover and market a new drug for a known target, with the probability of success starting at around 5%. This probability is even lower for a novel mechanism of disease. In stark contrast with the IT industry, where Moore’s law still applies, the pharmaceutical industry is setting new lows, aligning with Eroom’s law.
Several factors contribute to this lack of pharma productivity. The primary reason is the complexity of human biology and the limited tools available to study it safely. It has been challenging to turn drug discovery into an engineering problem. However, there is another rather unspoken trend contributing to decreased R&D productivity—increased rotation of C-level executives and R&D leaders in big pharma.
In my personal opinion (not the opinion of my company, Insilico Medicine), the biggest elephant in the pharma room is the new culture of frequent changes, where neither the executives nor the company are committed to specific therapeutic programs and are chasing new trends or career opportunities.
Strategic R&D restructuring
Many promising preclinical and clinical trial programs are terminated for reasons unrelated to science, drug efficacy, or safety. Each year, the pharmaceutical industry terminates hundreds, if not thousands, of programs for strategic reasons. These reasons range from cost-cutting, reprioritization of the sales force to facilitate newly acquired clinical-stage assets, or management changes at the top.
During my two decades in the biotechnology industry and nearly a decade at Insilico, I’ve witnessed the destructive power of these changes, resulting in massive losses of decades and hundreds of billions of dollars in research efforts. Since I founded Insilico, pharma companies such as GSK, Novartis, and Sanofi have changed R&D executive leadership twice and Pfizer, Roche, BMS, Bayer, and Merck KGaA once.
To my knowledge, over the past ten years, no AI-drug discovery company partnering with a big pharma company on target identification has completed a Phase I study or even reached Phase I. No company that I know of partnering on small-molecule design for a known target has reached Phase II. Hundreds, if not thousands of AI-pharma partnership announcements were made over the entire decade, but none have yet succeeded.
Like CEOs, CSOs and R&D heads are also changing frequently. For example, three talented young scientists changed corporate jobs relatively recently, two of whom did so three times within a short period. The turnover of R&D heads in pharma R&D in a number of pharmaceutical companies is higher than in some of the major tech companies.
The recent wave of pharma-AI partnerships at the platform level is unlikely to succeed either. At Insilico, on average we nominate a preclinical candidate for a moderately novel target in about a year. More novel targets may take up to 18 months internally. When partnering with a pharma company, we usually expect either a request for a novel target with few experimental assays available for validation or a “mission impossible” validated target that the company could not find a potent inhibitor for after years of trying. We usually never get the data from these years of failed experiments and need to start from scratch.
And the standards for preclinical candidate nomination are generally higher than would be set for a regular internal program. In both cases, it may take over 18 months to reach the preclinical candidate milestone. And within these 18 months, if the CEO, CSO, TA head, or even one or more of the key project personnel resigns or is let go for strategic purposes, the program may get deprioritized. When new management comes in, it likes to do new things.
Some of these organizational changes are just inconvenient and cause delays, but others can be deadly as those deprioritized drug programs may never be recovered due to patent issues and may even result in the collapse of the AI-powered drug discovery (AIDD) companies that were relying on them.
The bottom line is that the R&D leaders of big pharma companies are indeed like football coaches. They are almost always accomplished scientists or clinicians with impeccable credibility and scientific ambition. Once they come in, they restructure the team and change the strategy as they are expected to by their boards. The shareholders focus on shorter-term returns and want to see constant strategy adjustments, often disregarding the need for long-term program commitments.
These changes transpire every 4–5 years, while the game of drug discovery and development usually lasts 8–12 years when drugs are discovered from scratch and 6–7 years if the drugs are in-licensed at the preclinical candidate stage from a partner. The 4–5 year tenure is simply insufficient to see the internal or partnered program all the way to approval unless the program is acquired in later stages of development.
How can we reduce Big Pharma waste?
Unlike most other industries, the pharmaceutical industry impacts human lives in the most direct and profound way. The productivity of the pharmaceutical industry defines how long and how well each of us will live because, at the end of the day, we will all eventually be patients. When it comes to aging, all of us are already patients! Therefore, every leader in the pharmaceutical R&D industry is accountable not only to the shareholders but also to the reasonable expectations of society, who invest massively in the industry directly through taxes, insurance, and high costs of medicines (which also pays for R&D).
While it is apparent from the research studies on R&D productivity that the productivity in big pharma is low and a large part of this poor performance is the result of frequent strategy and personnel changes, I have not seen a single study trying to analyze the productivity of the individual CSOs and CEOs using a set of standardized metrics. For example, how many preclinical candidates they nominated using internal R&D, how many progressed into human clinical trials, and how many passed Phase II safety and early efficacy studies. There is also no metric evaluating the effectiveness of external partnerships.
![Considering the low probability of success of early-stage collaborations with pharma and relationship-driven decision-making, the best strategy the AI-powered drug discovery company may pursue is to commit to the discovery and development of the therapeutic program at least to Phase I completion and seek partnerships with pharma only when they reach the pre-clinical candidate stage or IND-enabling stage. [Insilico Medicine]](https://www.genengnews.com/wp-content/uploads/2024/09/7f5f482bct38e1546a35-2-e1726623360715-300x268.webp)
But in reality, it is easier than tracking the performance of scientists, where there are multiple standard metrics, including tools like the H-index and the number of publication citations. Such metrics would allow the pharmaceutical and biotechnology companies to be more transparent about their internal R&D performance and optimize it or develop new partnering capabilities.
The simple metrics I am proposing were outlined in a recent GEN article and are simple: number of preclinical candidates (PCCs) in a single year sourced internally and externally, the time it took to PCC and patent life remaining, and the number of PCCs progressed to Phase I, Phase II, Phase III along with the timelines. Such metrics would ensure that the companies could be compared using the common set of benchmarks in order to increase productivity of the entire industry. Otherwise, pharma companies should not even engage in internal R&D.
In the short term, large institutional investors that do care about environmental, societal, and governance (ESG) metrics should demand reports on per-program spending, program terminations, and plans for program recovery. If the program does not fit the pharmaceutical company strategy anymore, it should be outlicensed, given away to the other industry players, or openly published in peer-reviewed journals to ensure that the precious animal lives and scientists’ time is not wasted. Also, analysts following the pharma companies should demand such reports and provide comparisons using the standard set of metrics.
Otherwise, about 10–20% of company sales that is commonly re-invested into R&D on a regular basis is not only wasted but also toxic to the industry as many of the assets and targets that pharma abandons become unrecoverable or undevelopable due to the intellectual property issues or negative sentiment.
What can NextGen AI-powered drug discovery companies do better?
Considering the low probability of success of early-stage collaborations with pharma and relationship-driven decision-making, the best strategy the AI-powered drug discovery company may pursue is to commit to the discovery and development of the therapeutic program at least to Phase I completion and seek partnerships with pharma only when they reach the pre-clinical candidate stage or IND-enabling stage. This strategy would require investments of several million dollars per program to get to the preclinical candidate stage and tens of millions for Phase II to complete. It would also quickly demonstrate how effective AI approaches for target identification and validation compare with the more traditional approaches.
This approach will enable the AIDD companies to see their programs go into clinical trials and eventually reach the patients. In my opinion, the AIDD industry needs to favor long-term commitment and discourage serial entrepreneurs who continuously start companies, raise money, partner with pharma, and then leave to pursue other shiny opportunities. The starting goal should be the drug approval, not a test of a new algorithm or platform.
Prioritizing patients and programs over individual careers
In conclusion, the frequent organizational changes and restructuring in the big pharmaceutical companies create enormous waste. Every time the program is terminated for strategic and not scientific reasons, the many thousands of experimental animal lives, and the years of work of some of the smartest people on the planet are wasted. Moreover, when promising programs get terminated after patents on chemical scaffolds are filed, thousands of promising molecules are demonetized as no one would like to invest in the molecules patented by someone else or in-license if much of the patent life is wasted.
This culture of life-long commitment to programs was prevalent in the European and Japanese pharmaceutical firms, where executives often spent their entire careers in the same company and saw the programs they initiated make it to market. This culture has changed in favor of more dynamic executive, strategic, and R&D refreshes, a trend that may have contributed to the rapid increases of R&D costs. In contrast, executives and senior scientists of the world’s most valuable pharmaceutical company, Eli Lilly, have stayed with the company for over twenty years and managed to discover, develop, and put on the market several blockbusters targeting chronic diseases of aging.
Pharmaceutical industry executives today who change jobs in favor of a better title or pay package should realize that their career moves will inevitably result in this dramatic waste and reduce the probability of positive change in patients’ lives. All of us, biopharmaceutical executives and scientists alike, must be prepared to make personal sacrifices to see the therapeutic programs we start go through the finish line instead of chasing new and trendy career or scientific opportunities.
When we join this industry, getting more effective drugs for patients in need as quickly as possible should be the ultimate reward and objective. We also should remind ourselves that if we live long enough, we become patients in need.
Alex Zhavoronkov, PhD, is co-founder and co-CEO of Insilico Medicine. The opinions in this article are solely those of the author. Email: Alex Zhavoronkov, PhD <[email protected]>
