Clinical Reporting and Drug Discovery: A Glimpse of AI’s Impact on the Pandemic

The importance of AI in the pharmaceutical space has grown exponentially over the past few years, driven by global demand for a COVID vaccine at the height of the pandemic. Additionally, AI has proven useful throughout the pandemic, while offering a glimpse into the technology’s future promise.

According to a widely quoted 2021 Evaluation on the impact of AI applications in the wake of COVID-19 from the academic journal Frontiers in Medicinethese technologies have achieved “high performance in diagnosis, prognosis assessment, epidemic prediction and drug discovery for COVID-19”.

The report then concluded from their research that AI “has the potential to significantly improve the efficiency of the existing medical and healthcare system during the COVID-19 pandemic.”

In this article, we provide a brief overview of how AI has been used in the wake of the pandemic to:

• Accelerating drug discovery and production
• Streamline clinical reporting workflows
• Accelerate data validation

After a brief summary of the first two trends, we’ll take a closer look at a use case where Pfizer enlisted an AI vendor to help cleanse patient data to produce faster results for patients. clinical trials of their COVID-19 vaccine. Finally, we will conclude by taking a closer look at the frontiers of AI’s potential impact on clinical trial outcomes.

But first, let’s start our overview by taking a closer look at the impact of AI capabilities on the drug discovery process.

The future of drug discovery

The drug discovery process is extremely expensive and time-consuming. According to a report by PhRMA:

• The cycle from research to market takes at least ten years, with clinical trials taking an average of six to seven years.
• The average cost to develop an effective drug is $2.6 billion.
• The overall probability of clinical success (i.e. FDA approval) is estimated to be less than 12%.

The time spent in the research-to-market cycle for vaccines is specifically similaralthough less costly: up to $500 million, according to the International Vaccine Institute.

The pharmaceutical industry is under pressure to find new and better ways to manage the drug discovery process, and the answers are now more easily within reach. Big pharma – including Pfizer and Sanofi – use AI and machine learning to make this process faster and more cost effective.

In the case of Pfizer, the company managed to produce an FDA-approved vaccine in just under a year -; no small feat. According to a Pfizer Press release and substantiated by the media, AI and machine learning were used to accelerate the processing of a massive dataset to develop their PAXLOVID vaccine.

In a more recent case involving the development of an oral drug for COVID, Pfizer also claims to be using AI to help screen millions of potential compounds designed to affect molecular drug targets.

In the case of Sanofi, the pharmaceutical giant complaints on its website to use AI and machine learning to analyze anonymous data from around 450 million patients. The company recently published a press release revealing a partnership with Exscientia, an AI drug design and development company. The deal has a potential value of $5.2 billion.

Streamlined reporting process

Time is of the essence in the drug discovery process.

One of the most time-consuming parts of the drug discovery process is the preparation and generation of Clinical Study Reports (CSRs). According to a study Posted in Clinical researcherthe average time taken to complete the CSR is 109 days, or about 3.6 months.

This lengthy process not only uses company resources, but also prevents potentially useful drugs from advancing to potentially benefit patients.

Whatever the scenario, CSRs are essential; but much of the work is spent on repetitive tasks, requiring the valuable time of highly skilled healthcare professionals. AI can reduce the time taken by CSRs, freeing up the time of medical writers.

AI-enabled software tools can automate much of the CSR writing process. Time savings with CSR automation solutions vary by vendor. A company, ZYLiQ.ai, complaints to save medical writers 60-70% on time.

Another company, Narrativa, complaints a 65% reduction in time as well as an average cost reduction of 40% for CSR drafting.

Most user workflows of AI-based CSR solutions follow a similar pattern:

• Model selection and configuration (output)
• Download source documents
• System Generated CSR Review
• CSR User Edition
• Finalization and approval of release documents

Automating CSR content with AI helps simplify the review process by streamlining the generation of reliable, repeatable, and quality text. Among the range of current AI applications, natural language generation (NLG) is widely used in financial and medical writing for facilitate a consistent writing style throughout the document.

The CSR approach often requires information that is known only to a medical writer. Only a specific person may be able to explain the real value and meaning of the data.

In this case, the writer can program how the document will be read and how the facts fit together. NLP systems can be trained on the specific writing style and medical jargon required for a particular CSR.

A COVID Use Case: Accelerated Data Validation at Pfizer

After clinical trials, patient data needs to be “cleaned up” enough for scientists to accurately analyze the results. Moreover, the development of such a capacity would take a long time and could therefore further delay the process.

At the time, Pfizer was trying to get its vaccine approved for emergency use by the FDA and couldn’t afford any delays. The company decided to launch a competition to develop an AI-powered tool capable of quickly managing and cleaning clinical data.

A company called Saama Technologies won the competition with its Smart Data Query (SDQ) solution.

Saama states in his case study [pdf] report that the SDQ platform speeds up data cleansing and ensures data quality by automating the query management process. The report further states that Saama’s solution leverages the following AI capabilities:

• Machine learning to predict data deviations
• Natural language processing to help detect adverse event data and use medical history and case report forms for data consistency, and process over 750,000 free text sentences

Saama says the resulting AI-augmented workflow goes like this:

1. Site investigators feed electronic case report forms (eCRFs) into their electronic data capture (EDC) system, integrated with Saama’s Smart Data Query (SDQ) platform.
2. The SDQ platform reviews data and provides data managers with forecasts for their review.
3. The EDC automatically generates queries in the form of eCRFs, sorted by confidence intervals and highlighting any data discrepancies.
4. Managers then review each discrepancy and proposed response through the SDQ interface, assigning “open” requests to the eCRF already in the EDC or signing (step 6, shown in Figure 1 below). below) the correct changes.
5. SDQ recognizes any already correct query text and uses this data to improve its algorithm. If the query text contains errors, reviewers can edit the response (step 7) before issuing a query. SDQ applies the data from these corrections to its algorithm.

The report features a helpful illustration illustrating the ease of the above workflow (Figure 1 below):

(Figure 1. Source: Saama Technologies [pdf])

According to the case study report, the result was that the AI ​​reduced the median number of calendar days to generate queries from 25.4 days to 1.7 days across all vaccine studies.

Throughout the report, Samma also claims that their algorithms can help link potential treatments to precise biological causes in a much more efficient and rapid manner than the trial-and-error approach characterizing the traditional drug discovery process.

According to them, a lake of information can now be reduced to a small pool of relevant data in a significantly shortened time.

Looking Ahead: Accelerating Trial Outcomes

Another potential advantage of this new technology is that test results could be obtained much sooner. Driven by the Covid-19 outbreak, we are entering an era where algorithms can analyze clinical data and estimate simulated patient journeys through the trial, accurately predicting outcomes.

In a recent article for the famous academic journal Nature, it is reported that German biotech company Evotec was able to shorten the discovery process from 4-5 years to 8 months using AI.

By disrupting traditional testing methods with these capabilities, pharmaceutical companies can potentially reduce the average test cycle from years to hours. The AI ​​platform, ‘Centaur Chemist’ of exscientiacan supposedly “sort and compare various properties of millions of potential small molecules… for 10 or 20 to synthesize, test and optimize”.

Further investment and development can transform these AI-based methodologies into increasingly robust and reliable simulation tools for data-only clinical trials. A outstanding example of the current trend exists in the “Immune System GPS” model recently unveiled by Pfizer.