Download – Disruptive tech 2021 – 3 / 6 观点
A dozen years ago, the life sciences sector was largely dominated by small molecule pharmaceuticals; relatively simple chemical compounds that can be made by chemical synthesis. But this model, which had been successful for decades, was under pressure. In particular, using the methods available, it was getting harder to identify new targets for disease and match them with candidate drugs; meaning diminishing drug development pipelines.
At the same time, the key patents on a number of 'blockbuster' small molecule drugs that had driven the market in the previous decade were expiring, opening the market for generic competition and pushing market prices down.
How did businesses react? One trend that developed to fill drug pipelines was the acquisition of small research companies or licensing in their promising drug candidates. This had the advantage of avoiding much of the expensive and risky (most initial drug candidates fail) early-stage development in-house, and instead sharing this risk with the research company.
An alternative approach was to divest certain business operations, while retaining equity and intellectual property rights. This would produce smaller, more focused, teams dedicated to particular product goals, and had the advantage of removing the associated costs from the balance sheet.
In the same period, a marriage between the understanding of molecular biology and in silico technologies has enabled medical innovation at a seemingly unprecedented rate, driving the development of new business sectors. This has been borne out by the rate of patents filing in this area and the subject matter patent of disputes in the Patents Court. By these indicators the small molecule business – though still very important – is now matched by the biotechnology (biological products) and medical devices sectors.
Then there is the accelerant effect of the COVID-19 pandemic. This has resulted in hundreds of patent filings in 2020 and 2021 for remote medical technologies, as the medtech industry adapts to provide solutions for the socially distanced treatment of patients.
Artificial intelligence is an area of digital technology that is, with good reason, receiving much attention at the moment – the impact of AI in drug development is already becoming clear. Take, for example, the long-term collaboration formed between AstraZeneca and BenevolentAI. This is to use a combination of sophisticated algorithms and machine learning for the discovery and development of treatments for chronic kidney disease and idiopathic pulmonary fibrosis. And then, a similar collaboration between BenevolentAI and Novartis’s precision medicine team, this time to investigate further Novartis's existing oncology pipeline.
All these diseases are complex and the underlying disease biology is not well understood (thus 'idiopathic'). To investigate these diseases the interrogation of vast datasets is required. The work of both the AstraZeneca and Novartis collaborations is to combine known biomedical, genomic, clinical and molecular data, which has been extracted and contextualised using BenevolentAI’s proprietary 'knowledge graph', with a target identification platform. Using 'relation inference' models, possible disease targets can then be predicted and potential drug-like molecules generated with optimised treatment properties, ready for chemical synthesis.
In other words, computational techniques are being used to identify biological features, or ‘biomarkers’, that suggest certain drugs can be used for particular indications and/or subsets of patients within a particular disease group, and to work out what those drugs are.
The use of AI in combination with large datasets therefore promises to drive new and more targeted treatments. It is a trend that is not going to be reversed. Indeed, BenevolentAI claims that 90% of the world’s data was produced in the last two years alone.
A particularly important contributor to this process will be NHS England. As the largest of the UK's public health services, it has been accumulating patient data in various forms over 70 years. Once digitalised and anonymised, this will assist in identifying groups of people who are vulnerable to health risks and to predict which individuals are likely to benefit from healthcare interventions. The NHS also expects this data to be beneficial to companies within the life sciences sphere, as it makes it available to industry with the aim of driving research and innovation.
More data will come from what is claimed to be the world’s largest genome sequencing project. This has been funded by a consortium of UK Research and Innovation (a UK government agency), The Wellcome Trust, and four pharmaceutical companies (Amgen, AstraZeneca, GSK and Johnson & Johnson) and will sequence the genomes of 500,000 individuals at UK Biobank in Manchester.
There is still a lot of work and risk that lies between identification of a possible drug target and bringing a new treatment to market. But the ambition is that advances in molecular biology, digital technology and data collection, when fed into the appropriate AI systems will boost the finding of root causes of disease in specific, and smaller, groups of patients. It is also expected that the in silico methods offered by BenevolentAI and others will increasingly replace the costly and time-consuming traditional techniques, boosting drug discovery of small molecules and biologic drugs alike.
The rate of production of such data and the speed at which it can be structured, analysed and applied by AI will only increase in future. Drug companies may find that far from dwindling pipelines, they are spoilt for choice.
To learn more about pharmaceutical patents, check out A User's Guide to Intellectual Property in the Life Sciences, co-authored by Paul England.