Wither pharmaceutical innovation?

BenevolentAI- Scientist holding up Petri dish

Let’s talk pharmaceutical innovation, and in doing so let’s be honest and perhaps a little controversial. Last year the CEO of Regeneron, Leonard Schleifer, stood up at the Forbes Healthcare Summit in New York and said something surprisingly candid and revealing.

“The real reason we’re not liked, in my opinion, is because we as an industry have used price increases to cover up the gaps in innovation. That’s just a fact”.

Although his fellow panellists did not agree with him, this was in my view a profoundly accurate summary of the innovation vacuum that exists in pharma. There is clear evidence that this is the case. For example, the number of new drug approvals by the FDA in 2016 was 22 and of these, only 8 were first in class i.e. representing a drug tackling a new mechanism for the first time. Over the past ten years FDA approvals have averaged 29 per year with a low of 18 and a high of 45. For an industry that spends over $150 billion per annum on R&D that doesn’t seem to be a lot of innovation per $ spent.

One of the reasons for this lack of innovation can be found in data published in Nature Reviews Drug Discovery (2016 15:817–8) which showed that ~50% of compounds were terminated in Phase II and Phase III for lack of efficacy. Considering that costs increase exponentially in later phase drug development, this is a sobering statistic and indicates that more attention needs to be placed early on in development in selecting the best protein or gene, i.e. target in the body, to modulate therapeutically.

It might be surprising that we are still so poor as an industry at selecting the right targets to work on given the vast increase in biomedical data generation. But the sheer volume of this data, most of which is unstructured, makes it hard for researchers to efficiently mine that data for new connections.

These connections are needed not just for validating targets. Termination for safety reasons remain at 25% in Phase II and are still 14% in Phase III. These terminations are not split into those that were due to mechanistic toxicity and those that were due to compound based toxicity, but better mining of data could help to reduce both causes.

Cost reduction and improvements in success rates are absolutely essential if we are going to have a sustainable, innovative industry, that can deliver a pipeline of new medicines to treat serious diseases.

Drug development is lengthy — from identifying the optimal compound to getting it on the market can take up 15 years. As I mention at the top of this blog, it is also very very costly; with a new drug widely accepted as costing well over $1 billion to develop, and in fact recent studies suggest the figure is closer to $2 billion. Of course, this includes the cost of failures which is why even a relatively small improvement in success could be really important.

We believe here at BenevolentBio that AI and machine learning technology have the potential to change the way we do drug discovery — increasing the pace and the productivity of pharmaceutical organisations. How that can be achieved by both ourselves and others will be the subject of my next blog.

Jackie Hunter, CEO BenevolentBio



Uniting human and machine intelligence to discover new ways to treat disease | www.benevolent.com #Becauseitmatters #AI #DrugDiscovery #Innovation

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store

Uniting human and machine intelligence to discover new ways to treat disease | www.benevolent.com #Becauseitmatters #AI #DrugDiscovery #Innovation