Generative AI in Pharma: 5 ‘high-impact’ use cases for Gen AI in pharma today

There's no doubt that generative AI (gen AI) will profoundly transform the way biopharma companies develop, produce and market new treatments. R&D timelines have already been shortened considerably. Patients' quality of care and clinical outcomes are improving with AI-powered tools, while also becoming more cost-effective and affordable. Its potential across the entire value chain looks equally promising. 

The wide-ranging capabilities of gen AI raise questions about where and how to apply it, how fast to apply it, and how to manage the resulting large-scale changes.

Selecting and prioritizing the gen AI use cases that have the greatest impact on business value and cost reduction is paramount.  

To this end, The Boston Consulting Group (BCG) has identified 5 'high-impact' use cases for gen AI in the pharmaceutical market, with the potential to truly strengthen competitive positioning and create long-term advantage. 

It is estimated that gen AI solutions for the specific functions as identified could boost productivity by up to 30%.

What are the most promising use cases for gen AI in pharma today?

The 5 ‘high-impact’ use cases for gen AI identified by the BCG are as follows: 

·      Faster drug molecular design and discovery.

Gen AI can accelerate early-stage drug breakthroughs by assisting in the discovery and optimization of small- and large-molecule drug candidates. 

Gen AI is a real game-changer for the identification of potential new drugs. It is used to create new molecular structures targeting precisely identified conditions, paving the way for new compounds to fight previously untreatable diseases. By generating drugs optimized for efficacy and safety, gen AI expedites the traditionally lengthy drug discovery process.

When a promising drug candidate is identified, Gen AI helps refine its structure. This process, known as lead optimization, adjusts the molecular structure to improve the drug's pharmacological properties. 

Gen AI has been successfully used in this context for many years, reducing lead times by 25%. Gen AI can be expected to accelerate drug design further.

·      Accelerated clinical development and access. 

Gen AI optimizes and accelerates clinical development and access to medicines in three principal ways: 

Firstly, gen AI improves clinical trial design and helps manage trial performance by generating simpler, more efficient protocols and using simulations to prevent delays. By analyzing electronic medical records (EMRs), AI identifies appropriate patient types. This targeted approach improves trial success rates by focusing on those most likely to respond to treatment. AI's capabilities extend to research design. Through the use of ‘digital twins’, it simulates placebo cohorts, reducing the size of control groups. In addition, AI streamlines the processing of the large amount of data involved in trials, facilitating study organization and expediting the consent process.

Secondly, gen AI automates the generation of medical documents in clinical development (e.g. protocols, clinical study reports and regulatory affairs submissions), cutting down medical writing time by as much as 30%. AI-driven text summarization tools analyze large amounts of data to extract key information. This speeds up analysis and supports informed decision-making. 

Thirdly, gen AI will help the development of decentralized clinical trials (DCTs). For those unfamiliar with DCTs, these are trials in which some or all of the clinical trial activities occur at locations other than a traditional clinical trial site. These alternate locations may include the participant's home, a local healthcare facility, or a nearby laboratory that integrates digital health technologies (DHTs), which are systems that capture health information related to clinical trials directly from individuals.

According to Dr. Elzharrad, FDA: "Decentralized designs and the use of digital health technology (DHT) offer great potential for streamlining clinical trials in general, but also for expanding the scope of trials and easing the burden on participants. Many such DHTs are AI-enabled, whether in the form of algorithms embedded in the DHT itself, or by exploiting its data after collection." 

One of FDA's key priorities is to advance the modernization of clinical trials, seen as the cornerstone of evidence generation. Specifically, AI-powered algorithms are capable of detecting clusters of signs and symptoms to identify potential safety signals – in real time – one of AI's most powerful capabilities.

From 2016 to the present, around 300 applications filed with the FDA have reported the use of AI, according to the FDA's Dr. Elzharrad.

·      More efficient quality management and submission. 

As part of the quality assurance process in manufacturing, Gen AI can augment or automate the analysis and generation of documents in support of quality and regulatory compliance, for example by managing deviations from standard procedures and supporting annual product quality reviews.

The technology can highlight recurring patterns and issues in large volumes of documentation from a variety of sources (such as quality systems, manufacturing systems and supplier data) and help generate preliminary reports on quality outcomes that are essential for regulatory compliance. 

BCG estimates that streamlining these routine tasks could boost staff productivity by 30%, thereby expediting the drug approval process.

·      More targeted content creation, tailored for personalized interactions

Gen AI helps collect and organize complex information from a variety of data sources to create truly personalized interactions – in support of frontline sales and medical teams, for example. 

The challenge is to interpret large quantities of ever-changing data to establish clear priorities and create personalized messages for a myriad of different types of healthcare professionals. 

For example, to improve targeting and promotion decisions and guide sales reps' actions more precisely, AI can help inform them of whom to call on and in what way. For example, an oncology drug should be targeted by physician specialty, geography, typical ages seen, conditions typically treated, the mix of insurance used, patient numbers, institutional affiliation, and whether the physician is being seen for first, second-, or third-line treatment. There are dozens or hundreds of potential combinations of these factors.

As drugs become ever more targeted, the market for each drug becomes smaller and the value of a single patient grows. With each patient potentially worth tens of thousands of dollars – or more – this targeting is essential. At the same time, AI may enable trimming expenses on sales reps by guiding their actions more directly, and potentially initiating automated digital actions instead.

Gen AI can also adapt content to different cultural contexts and channels. Applications include training, supporting interactions with healthcare professionals, and streamlining patient support services. 

BCG expects such efforts to generate revenue increases of up to 10% and to reduce external agency costs by 25% or more.

·      Facilitated review process.

As content becomes increasingly personalized, medical, legal and regulatory reviews will become an ever-bigger bottleneck. Gen AI can facilitate the review process by automating the pre-screening of materials and assisting in the parts of the process performed by human staff (identifying risks, referencing claims and fine-tuning drafts, for example). 

According to BCG, gen AI can increase the productivity of these high-intensity tasks by up to 40%.

Final thoughts

Historically, only rare innovations have enabled simultaneous leaps upward in value and downwards in cost. The use of AI could begin to make such innovations commonplace.

From the identification of target patients using unstructured data from medical records to the generation of personalized promotional and medical texts, all the way to interactive digital sales agents and self-managed customer support, generative AI will lead to revolutionary innovations in the pharmaceutical sector.

However, implementing gen AI is tricky in a complex industry that is subject to intense regulatory scrutiny. The wide-ranging capabilities of gen AI raises challenges about where and how to apply it, how fast to apply it, and how to manage the resulting large-scale changes. 

To create the conditions for testing and expanding these applications, technology and strategy need to converge, so that the right choices can be made for each organization's specific circumstances.