AI and the pharmaceutical industry: a slow but sure win-win combination

The pharmaceutical industry has long invested in technology and innovations aiming to reduce costs and time for drug discovery and development. Being one of the few industries where bringing a product to market takes approximately a decade and billions of Euros of investment, it is natural that pharmaceutical companies are in a continuous lookout for cost-cutting innovations. 

AI and the pharmaceutical industry: a slow but sure win-win combination

With enormous amounts of siloed health data generated, and complex regulatory environments the pharma industry is ripe to benefit from the adoption of AI strategies. Yet, the pharmaceutical industry has been slower than other industries in leveraging the potential of AI. A survey conducted by Tata Consulting and BitKom specifies that only 43% of chemical and pharmaceutical companies in Germany are even open to artificial intelligence. Businesses across all industries cite data protection and data security amongst the top reasons for the reluctance to invest in AI. Globally however, according to a 2019 Emerj research of more than 50 executives of healthcare companies, it is believed that there will be a broad scale adoption of AI in healthcare by 2025. With 50% of the research respondents believing that AI will be ubiquitous in healthcare in 2025.

Collaborations between pharmaceutical companies and AI 

Not surprisingly big pharma companies have already started to collaborate and invest in AI. LEK consulting conducted a comprehensive research in 2018 (Figure 1), which clearly shows that the pharmaceutical industry can apply AI in all phases of the drug lifecycle.  

Figure 1 depicts an interesting picture, where the highest distribution of AI use cases is initially concentrated in the drug discovery phase. This reflects the lower regulatory implications of this phase compared to clinical trials. However, AI developments already in 2018 seemed to be present across the entire spectrum of the drug lifecycle and pharma business from target discovery to post-approval activities.  

The LEK research also showed that pharma companies were already collaborating with more than one AI company to leverage AI advantages. Roche partnered with Owkin to speed up drug discovery and development, acquired Flatiron to accelerate cancer research, and GNS Healthcare to use machine learning to improve patient care in oncology. Another company Roche recently acquired is Viewics, Inc, which focuses on business analytics for laboratories, to help improve the efficiency of the lab. Novartis on the other hand, after collaborating with MIT, IBM Watson, Intel and QUANTUMBLACK, in 2019 also announced a collaboration with Microsoft. The collaboration was set to transform medicine with artificial intelligence and establishthe Novartis AI Innovation Lab. The aim of the Innovation Lab is to bolster Novartis AI capabilities from research through commercialization and help accelerate the discovery and development of transformative medicines for patients worldwide.

The shopping / collaboration spree for / with AI companies continued. According to the Guardian 90% of the big pharma companies, initiated AI projects in 2020. And biopharmatrend highlights a significant number of AI focused collaborations of "big pharma" players as well as the fact that the sector is heating up for investments. Drug discovery continues to go strong with Bayer entering a $ 266 million agreement with Excentia to accelerate drug discovery programs in oncology and cardiology. Soon after Bayer signs a five-year partnership with Schroedinger to screen and design synthetically feasible compounds. Pfizer teamed up with Insilico Medicine planning to identify new targets and biomarkers.Two months later, Pfizer announced that it will collaborate with Saama to improve clinical trial models and accelerate getting drugs to the market. Boehringer Ingelheim signed deals with Insilico Medicine and Berg Health. Roche Group subsidiary Genentech selected a start-up drug discovery firm Genesis Therapeutics for drug discovery. 

Clearly, across the industry collaborations and acquisitions are mushrooming, paving the way for promising developments and innovations in the adoption of AI for pharma. 

The Infrastructure AI Platforms and Pharma

An example of an early collaboration which is now reaping benefits is the collaboration announced in 2017 between two giants in their respective industries Merck Group and Palantir technologies. It clearly highlighted how Big Data analysts would trawl through Merck's data stockpile to guide the development of new drugs and therapies, with Merck CEO stating that the sky is the limit in terms of what they could achieve together. Indeed, a year later Syntropy was born. A joint venture of the two, Syntropy facilitates collaboration within the global scientific community, in order to drive breakthrough innovation in cancer research.The proposed data integration platform leverages Palantir Foundry - an operating system for the enterprise claiming to have several potential applications - which in this case allows users to structure and analyze data from multiple sources. As a result of this collaboration and Palantir's increased expertise in pharma it won in 2020 a contract with the FDA to power drug reviews and inspections.

This particular case is worth setting apart as among the challenges to adopt AI healthcare weekly lists the unfamiliarity of the technology, lack of proper IT infrastructure and the inappropriate data formats. AI infrastructure platforms provide end-to-end solutions which unlock the value of data trapped in silos by providing the necessary infrastructure and most importantly being able to work with various data formats. A myriad of applications can then be implemented on the existing infrastructure. 

Partnering with companies that offer end-to-end platform solutions for AI development and application, can prove a recipe for success for pharma companies. Various start-ups are now leveraging new technologies to tap into the value of healthcare data. AICURA medical is another example. The Berlin based start-up promises to industrialize lifecycle management of data in the healthcare industry across heterogenous data silos. Born in the land of GDPR, AICURA prides itself for its high standards of data protection and security. 

The company is already working with the largest hospitals and research organizations in Germany to streamline several AI use cases. And is in discussion with several pharma and CRO companies to tailor new use cases or increase infrastructural efficiencies in implementing existing ones. 

The journey has been slow 

The journey of AI adoption in pharma has been slow. In 2013 pharma companies were reluctant to invest in AI since there were few examples of pharmaceutical companies creating value from the improved use of big data, according to McKinsey . In 2018 most of the big pharma companies had already dipped their toes in AI and were collaborating with more than one AI company to build a number of use cases that show the impact and benefits of using AI across their business value chains. In 2020 it became clear that collaborations were the best way to adopt AI in pharma with more than 90% of the big pharmaceutical companies involved in AI projects. And according to biopharmatrendthese included partnerships on a variety of topics such as small molecules identification, drug repurposing, clinical trials and precision medicine, target identification,

It would seem like the industry is quite advance at this stage. However, a reputable study (called by Forbes the one study shaking the pharmaceutical industry) that conducts a comprehensive overview and comparison of pharmaceutical companies doing AI, concluded that “the pharma industry is in an 'early mature' phase of using AI in R&D. And furthermore, they could demonstrate that regardless of the efforts that need to be managed, recent developments in the industry indicate that it is worthwhile to invest to become a 'digital pharma player'. " Providing scientific proof that there is still so much value to be generated by leveraging big data and AI in pharma. 

But the future looks bright

With the now recognized potential of AI applications in the pharmaceutical industry and the actual use of AI being in an early mature phase, the future looks bright for AI and pharma.  

Global Data depicts AI as the most disruptive technology across the pharmaceutical industry in 2021 and beyond, despite the initial industry's slow adoption of the technology. 

And with already several successful collaborations, winning use cases, and scientific proof that investing to become a digital player is the right way forward, pharma will continue to invest in AI companies, and collaborations.

Building on AI-powered organization is a long process that depends on culture just as much as on technology, and it is quiet not sure which of the pharma companies will become one. Starting off with projects and betting on start-ups is the right way to go. However, getting stuck in “ pilot purgatory ” is a real threat and many companies do not manage to scale AI projects across the entire organization. Investing early in infrastructure platforms like Palantir Foundry or AICURA Platform, that allow for many use cases to be developed at a fraction of the time and cost, which leads to the ability to easily scale AI, may make all the difference. 
 

Edlira Kasaj, Head of Marketing and Commercial, May 2021