Contributed Commentary By Kenneth L. Massey and Stephen Cunningham

July 2, 2019 | Within the life sciences industry, there is general recognition that data analytics can streamline the drug research and development processes. But given the breadth, depth, scope, and challenges across the R&D continuum, the biopharmaceutical industry may not have clear line of sight to the comprehensive benefits of AI-powered data ingestion and harmonization platforms and the advantages such intuitive human-to-computer interface brings to drug development. To facilitate that understanding, here is an overview of the top ten R&D pain points that state-of-the-art analytics solutions can solve for sponsors, both in the research and the development phases of a program's lifecycle.

1. Pinpointing Biological Targets

AI-powered analytics platforms offer tremendous promise and opportunity to quickly and effectively integrate early biological and genomic data to identify promising drug targets. These solutions leverage proprietary machine learning algorithms to ingest enormous amounts of disparate biological data and pinpoint potential therapeutic targets that might have previously been overlooked.

2. Identifying Potential Leads

Identifying potential leads to address a biological target involves massive amounts of information. Traditionally, this is a resource-intensive and time-consuming process as researchers integrate biology, medicinal chemistry, and pharmacology data to develop 3-D molecular views and address critical issues, such as identification of potential receptors, molecule-target interaction profile, and the potential for off-target effects. Advances in data-driven methodologies enable these assessments to be made more accurately and rapidly, minimizing error and saving time and money.

3. Optimizing the Selected Lead

Lead optimization is a very data-intensive phase of research, involving the evaluation and optimization of the compound in terms of its biological activity, pharmacokinetics, pharmacodynamics, and potential toxicity profile. Data analytics eases the computational strain and rigor inherent in this phase, and helps researchers visualize the information as ideally as possible.

4. Broadening the Safety Data

As scientists start testing in animal models and move into the clinic, the early collation and interpretation of safety data and safety signals becomes paramount. AI models can effectively process the entire safety database almost instantly, allowing clinicians to understand the patient's condition and response to medications. This facilitates an early understanding of potential safety signals, and early and accurate dialogue with regulatory agencies. Regulatory agency partnerships based on shared databases can greatly enhance the development of new medicines.

5. Patient Identification & Recruitment

During study planning, data analytics solutions empower researchers to leverage real world data to optimize protocol design by evaluating the impact various inclusion-exclusion criteria have on the number of eligible patients, as well as disease prevalence and distribution to identify "hot spots" of potential subject density. Thoughtful evaluation of these criteria helps avoid unnecessary screen failures and recruitment of subjects who ultimately do not complete the study.

6. Site Selection & Management

According to clinicaltrials.gov, over 11% of trial sites fail to enroll a single patient, and almost 40% miss enrollment targets. Sponsors cannot afford these statistics. Platforms augmented by machine learning aid in site selection based on predicted performance for a trial's phase and therapeutic area. These solutions also enable investigators to assess whether sites are meeting performance targets, and compare study performance across and between sites. Quickly gaining insights into anomalies and discrepancies early with data analytics allows the team to make quicker decisions and midcourse corrections to get a study back on track.

7. Investigator Selection

Recruitment of the right principle investigators is critical to the success of a clinical trial. Advanced data analytics capabilities can help identify eligible PI populations and recommend those that best complement the needs of a given trial and therapeutic area. Patient flow analysis, size and scope of the referral network, and impact of competing trials can all be informed with AI platforms.

8. Global Portfolio Analysis

Clinical study conduct and management involves multiple stakeholders. Trial, regional, and country managers each require a different operational view to make decisions according to their roles. To keep a study on track and budget, each stakeholder's access to the right data and analysis is essential. AI data analytics platforms help sponsors track and monitor one or an entire portfolio of studies, empowering stakeholders to manage their assigned studies.

9. Operational and Financial Risk Management & Mitigation

A typical clinical trial management system compares the pre-initiation plan with the actual trial status, indicating discrepancies. It's not sophisticated enough to anticipate red flags, nor can it leverage learnings from similar past trials to forecast future KPI achievement. AI-powered platforms with pre-trained algorithms provide sponsors with the ability to predict operational and financial KPI status, flag operational or financial deviations from business objectives, and take immediate remedial action.

10. Data Quality & Compliance

With increasingly complex protocols and ever-tightening regulations, sponsors are adopting AI-powered analytics platforms that monitor study protocols and processes with higher accuracy and wider reach than traditional site visits, audits and inspections. These platforms ensure data quality across all recording systems, while maintaining compliance standards. Such solutions also ensure the representation of a single source of truth and track periodic data entry lags, adverse events, protocol deviations, types of issues, and resolution rates, making the data available for inspection or analysis in real-time.

Clinical and pre-clinical data are recognized to be key corporate assets that provide critical evidence of a medicine's efficacy, safety and potential economic value in the market. Clinical development has fallen behind in adopting digital technologies, which change how organizations engage patients, innovate in patient care and extract efficiencies in the development of new medicines. Data ingestion and harmonization platforms create an architecture of intuitive, AI-related algorithms and interfaces that standardize, interpret and present information in real time. Such solutions facilitate better decision making, improve quality, minimize costs, reduce sponsors' developmental timelines, and, importantly, help deliver needed therapies to patients.

Dr. Stephen Cunningham is a 25-year pharmaceutical veteran who has led teams that delivered on the development, regulatory approval and launch of numerous therapeutics. He has global experience leading cross-functional teams from Phase 1 thru life cycle management in neuroscience, cardiovascular, metabolism, respiratory, immunology, oncology, women’s health, transplant and clinical pharmacology. Dr. Cunningham is a member of Saama's Clinical Board of Advisors. He can be reached at lurgan@optonline.net.

Dr. Kenneth L. Massey is Chief Life Sciences Officer at Saama Technologies, responsible for driving market leadership for Saama's award-winning life sciences data analytics solutions. Prior to joining Saama, Dr. Massey had a 25-year career in the pharmaceutical industry, successfully leading and growing teams in clinical operations and global medical affairs. He can be reached at ken.massey@saama.com.