Insights

September 5, 2023
As AI permeates business functions across nearly every industry, communicators can glean important lessons from the way each sector talks about the technological advancement and disruption in their respective fields. 

With respect to healthcare, researchers have leveraged AI in medicine for years, and we are beginning to see how life-changing treatments can reach the market much faster. 

• The “patent cliff” – when the world’s 10 biggest drugmakers stand to lose nearly half their revenue by the end of the decade – is fast approaching. 
• Meanwhile, more than 150 small-molecule drugs are in discovery using an AI-first approach, with more than 15 in clinical trials. The annual growth rate for that pipeline is nearly 40%, according to the Boston Consulting Group.

In this issue, we explore the evolving use of AI in drug discovery, and with it, the rising potential of real-word evidence (RWE). 

Then, we’ll evaluate the essential role that communicators play in shaping public perception and dialogue around the use of AI in drug and medical device development.

1. Moving from concept to market faster. How AI creates efficiencies in drug discovery

The big picture: Estimates vary, but it currently costs about $1 billion and takes roughly 10 years to develop a new drug, with only a fraction of them making it to the market. 

• Change won’t be immediate. But AI can help scientists discover a drug faster by predicting how different molecules might behave in the body, and discarding dead-end compounds so promising candidates make it to clinical trials quicker.
• While there is no shortcut in human clinical trials, AI can optimize and diversify patient pools by identifying high-potential candidates. Currently, just 5% of eligible patients participate in clinical research, which limits the ability to study drug efficacy for specific subgroups.

Go deeper: Decentralized clinical trials can facilitate patient engagement by using remote monitoring via wearable devices, which transmit real-world data (RWD) like vital signs and medication adherence to researchers. 

• Researchers can use AI to analyze RWD for potential adverse events and safety signals, allowing earlier detection of potential drug safety issues.
• In some cases, AI is helping drug companies bypass the animal testing stage, allowing them to use computer models of humans instead. Machine learning can also accelerate the repurposing of existing drugs, which is patentable.

What else? Rare diseases get a leg up from the Orphan Drug Tax Credit and the FDA’s fast track designation, but their small patient pools present tough challenges that discourage drugmakers from prioritizing research in this space. 

• As a result, 95% of rare diseases have no approved treatments.
• AI is getting better at finding subtle links in large swaths of information that even the finest minds could miss, which helps researchers repurpose drugs and develop new ones faster, even without a large sample size.

What they’re saying:

• Eric Topol, Scripps Research Translational Institute: “There is no shortage of interest [in AI]. Every major pharma company has invested in partnerships with at least one, if not multiple, AI companies.”

• David Ricks, Eli Lilly: “In a discovery process, you want to funnel wide. In the past, perhaps humans would just think of what they already knew about. The machine doesn’t. It just knows about everything that was there and it comes up with constructs that humans just don’t.” 

• Tim Guilliams, Healx: “The potential to suddenly create a viable pipeline for many conditions with only a handful of patients, at the very least, gives real hope.”

Yes, but: Jim Weatherall, AstraZeneca’s VP of data science, AI and R&D, said the challenge for the next few years is pull-through, or to actually bring these drugs to market. He is otherwise optimistic: “We’ve been on a journey from ‘what is this?’ to ‘why did we ever do it any other way?’”

2. AI bolsters the pipeline from RWD to RWE

The big picture: Successful AI drug development requires high-quality, real world data, which is challenging to obtain and can be rife with privacy implications. RWD often comprises electronic health records, which present challenges at scale due to a lack of standardization (as they are collected outside the controlled environment of a clinical trial).

• Some researchers believe the answer to these concerns could lie in synthetic data produced by applying predictive AI algorithms to RWD. In pharma, synthetic data could be used to handle large but sensitive samples, where regulatory restrictions and data privacy are involved, such as in cross-border research.
• For now, synthetic data is a niche pursuit and hasn’t yet made its way into clinical use, largely due to concerns that it inaccurately represents the target population.

Go deeper: “The complexity and the variability in healthcare and science makes it a really hard problem to solve,” said Jim Swanson, chief information officer of Johnson & Johnson. “You can create synthetic data easily enough, but is it correlated enough to give you a specific and accurate example? That’s the problem you have to solve.”

• As such, RWD is used increasingly throughout the drug development process, from identifying early targets to post-market safety surveillance. 
• The ability to convert RWD to RWE using analytics is a crucial measure of success, as regulators recognize the benefits of RWE and fold them into decision-making. 
• This is where AI comes in. Algorithms can identify patterns and relationships within RWD to produce RWE. It can then be used to predict patient outcomes and compare treatments to help researchers understand which are more effective and safe in the real-world setting.

3. Evaluating implications for communicators

The biopharma industry is on a precipice. A Morgan Stanley report estimates that even a modest improvement in early-stage drug development success rates could bring 50 novel therapies to market over 10 years.

After discovery comes the story. 

• Communicating new science is tricky and can have a lasting negative impact if not done right. 
• The challenge is figuring out how to communicate AI’s benefits and ethical considerations in medicine – when the first AI-developed drug eventually hits the market.

Here are five key considerations for communicators.

1. Understand and be transparent about AI’s capabilities and limitations to build trust. Don’t shy away from the risks. 
2. Be authentic and clear about the potential and limitations of AI. 
3. Be true to the work and its impact. Use data and insights to educate. Leverage publications and medical meetings as opportunities. 
4. Showcase the significant personal and societal impact of healthcare innovation on patients over the last century, with AI as the latest example.
5. Proactively address concerns about data privacy and AI biases. Clearly communicate how your AI solutions adhere to regulations and best practices. Consider working with medical experts to create a campaign that speaks to the worries and anxieties of the public.

Share your feedback with us at insights@gga.nyc.