Top 10 ways for pharma to leverage AI according to... AI
![AI in pharma](/sites/default/files/styles/x_large/public/2024-05/Circuit-Board-AI.jpg?itok=tr0gh6WA)
Like most people these days, I have been experimenting with various artificial intelligence (AI) tools this past year. It’s remarkable to see the progress made in the technology compared to only a few months ago and there are no signs of the progress slowing down.
The pharmaceutical industry has numerous use cases for AI. From the drug discovery phase to patient care, AI has the potential to enhance almost every stage of the product life cycle. Out of curiosity, I asked a generative AI (GenAI) model what it considers the top innovative and new ways that AI tools can be employed to enhance pharma’s operations. With some paraphrasing, here’s what it said:
- Identification of biological data for drug discovery and development
- Analysis of patient data, including genetic information, to help develop personalised medicine
- Optimisation of clinical trial design and execution
- Improved manufacturing efficiency
- Enhanced supply chain visibility and forecasting
- Analysis of real-world data to inform better decision-making
- Digital therapeutics and patient monitoring
- Help with navigating the complex regulatory landscape
- Market analysis and competitive intelligence
- Drug repurposing
Let’s dive in!
1. Drug discovery and development
In theory, AI can substantially reduce the time and cost of drug discovery and development in four main ways: access to new biology, improved or novel chemistry, better success rates, and quicker and cheaper discovery processes. By predicting how different compounds will react with each other and with their targets, and by quickly analysing vast amounts of biological data, AI can identify drug candidates that are more likely to succeed in clinical trials.
Unsurprisingly, the US FDA has reported a noticeable increase in the number of submissions that use AI/machine learning (ML) components in recent years; in 2021, they received more than 100 submissions wherein AI use was reported. In 2022, the AI-fuelled pipeline reportedly expanded at an annual rate of almost 40%. Looking back at 2023, AI for drug discovery saw both wins and losses. Despite new partnerships and increasing traction, there were some setbacks. Notably, the AI-aided drug ulotaront, a TAAR1 agonist developed for the treatment of schizophrenia, failed its two Phase 3 studies.
2. Personalised medicine
AI can help develop precision medicine by analysing patient data, including demographic, clinical, and genetic information, as well as information about their metabolism and microbes, environmental exposures, and lifestyle factors, to predict how different patients will respond to various therapies. In theory, this will allow for highly tailored, personalised therapy plans with superior efficacy and minimal side effects.
Beyond pharmacogenomics, AI’s role in personalised medicine expands to precision dosing and disease prediction/prevention, including AI-powered risk assessment and lifestyle recommendations.
3. Clinical trials
Clinical trials are riddled with issues and are known to be highly time- and cost-intensive. AI has the potential to optimise clinical trial design and execution in multiple ways. Among others, ML algorithms can predict trial outcomes, write protocols, optimise eligibility criteria, identify and recruit the most suitable candidates for trials, monitor patient health remotely, and analyse data. When used to augment human-centric efforts, such as virtual patient and provider insight-gathering, AI tools could help recruit and retain more diverse participants, improve the efficiency and effectiveness of clinical trials, lower the associated costs, and expedite the drug development process.
4. Manufacturing optimisation
While the manufacturing industry lags behind others in terms of AI uptake, there is no shortage of potential ways to leverage this technology. AI can be used to improve the efficiency of pharmaceutical manufacturing processes through ongoing monitoring and predictive maintenance of equipment, optimisation of production processes, and automated quality control, helping to reduce costs and improve yield. It also has the potential to reduce the high carbon footprint associated with manufacturing.
5. Supply chain management
Due to geopolitical and climate-related challenges, pharma supply chain disruptions are becoming increasingly common and complex. AI algorithms can analyse extensive datasets in real-time to efficiently and dynamically forecast demand fluctuations, optimise production schedules, and proactively avoid bottlenecks in the pharma supply chain. In turn, this can support decision-making and lead to more efficient inventory management and faster delivery times, enhance supply chain visibility and transparency, and avoid overproduction and waste.
Interestingly, a 2023 study found that, in emerging markets, technological factors, including technology infrastructure and feasibility to implement or adopt AI at an organisational level, are the most influential factors impacting AI adoption in the healthcare supply chain, followed by institutional or environmental, human, and organisational dimensions
6. Real-world evidence (RWE)
Real-word data from electronic health records, insurance claims, wearables, social media, and other sources can provide important insights into real-life drug performance, patient adherence, and market trends to inform better decision-making. In 2020, McKinsey estimated that over the next three to five years, the average Top 20 Big Pharma company could unlock over $300 million a year by adopting advanced RWE analytics across its value chain. This technology could help identify new drug targets, shorten the time to market, improve formulary position and payer negotiations, and generate stronger evidence of differentiation and benefit/risk balance for in-market products. However, despite the great promise, many companies still struggle to deploy AI and advanced analytics effectively to generate RWE.
7. Digital therapeutics and patient monitoring
Patient health, behavioural, and outcome data from digital therapeutics and remote patient monitoring, including wearable devices, can be effectively analysed using advanced analytics and AI to identify patterns, trends, and predictive insights. As a result, AI-powered monitoring tools can provide patients with personalised treatment recommendations, while continuously monitoring their health and alerting healthcare providers to potential issues in real-time. Studies have shown that this has the potential to not only improve clinical outcomes, reduce the incidence of complications, and shorten hospital stays, but also to improve patient satisfaction.
8. Regulatory compliance
Regulatory frameworks are continuously evolving in the pharmaceutical industry. To avoid compliance issues, it is essential to keep up with new regulatory changes, including updated reporting guidelines and safety requirements. AI and Natural Language Processing (NLP) tools can help pharma companies navigate the increasingly complex regulatory landscape through regulatory intelligence and by automating the compilation and review of regulatory documents, predicting regulatory risks, and ensuring compliance with global standards.
9. Market analysis and competitive intelligence
AI tools can analyse vast amounts of data from various sources to provide insights into market trends, competitor activities, and emerging opportunities. By automating these tedious manual assessments, AI can substantially widen the scope of evaluation and help pharma companies make more informed strategic decisions and optimise their drug launches.
10. Drug repurposing
Lastly, AI could play a role in drug repurposing, i.e., the use of approved or investigational drugs to treat conditions other than the one they were initially developed for. Drug repurposing has the benefit of delivering safe and effective therapeutics to patients much faster and cost-efficiently than a de novo therapeutic, and AI can speed this process up even more by identifying new and promising targets. Specifically, AI can help with data mining and integration, predictive modelling and virtual screening, network pharmacology and systems biology, clinical trial optimisation, safety and toxicity prediction, and decision support.
Importance of the human touch
As AI technology continues to evolve, its applications in the pharmaceutical industry are likely to expand, offering new opportunities to improve health outcomes and transform business operations. However, while great progress has been made lately, it is important to keep in mind that AI should only supplement, never replace, human efforts. Just like asking AI to help me outline and find references for (but not write) this article saved me countless hours that can now be used on something else, pharma can leverage the vast array of AI tools to enhance their own productivity and optimise their operations.